View metadata, citation and similar papers at core.ac.uk brought to you by CORE
provided by University of Birmingham Research Portal
Chapter Divisions, Capitula Lists, and the Old Latin Versions of John Houghton, H.A.G.
DOI: 10.1484/J.RB.5.100457 License: Creative Commons: Attribution (CC BY)
Document Version Publisher's PDF, also known as Version of record Citation for published version (Harvard): Houghton, H 2011, 'Chapter Divisions, Capitula Lists, and the Old Latin Versions of John', Revue Bénédictine, vol. 121, no. 2, pp. 316-356. https://doi.org/10.1484/J.RB.5.100457
Link to publication on Research at Birmingham portal
Publisher Rights Statement: Eligibility for repository : checked 17/07/2015
General rights Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes permitted by law.
• Users may freely distribute the URL that is used to identify this publication. • Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. • User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) • Users may not further distribute the material nor use it for the purposes of commercial gain.
Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.
When citing, please reference the published version. Take down policy While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been uploaded in error or has been deemed to be commercially or otherwise sensitive.
If you believe that this is the case for this document, please contact [email protected] providing details and we will remove access to the work immediately and investigate.
Download date: 01. Feb. 2019 CHAPTER DIVISIONS, CAPITULA LISTS, AND THE OLD LATIN VERSIONS OF JOHN
Introduction* The modern division of the Gospel according to John into twenty- one chapters dates back only as far as Stephen Langton around the year 1200, while the current system of verses is largely the work of the printer Stephanus (Robert Étienne) in the sixteenth century.1 Biblical manuscripts present a number of different series. In the Greek tradition, the most widespread type is one of eighteen chapters found in the fifth- century Codex Alexandrinus (Gregory-Aland 02) and numerous subse- quent Byzantine manuscripts. These are known as kephalaia, and a list of chapter titles usually precedes the text of the Gospel.2 An earlier but very rare alternative is the set of eighty numbered paragraphs in Codex Vaticanus (G-A 03).3 Smaller divisions of text are supplied by the Euse-
* At the time of writing, the author was a Research Fellow on the Vetus Latina Iohannes Project, funded by the Arts and Humanities Research Council. He would like to thank Pierre-Maurice Bogaert and Patrick McGurk for comments on earlier drafts of this article, and Jeffrey J. Kloha for verifying readings in Vatican, Reg. lat. 14. 1. On the introduction of these systems, see Jean Vezin, « Les divisions du texte dans les Évangiles jusqu’à l’apparition de l’imprimerie » in Alfonso Maieru ` (ed.), Grafia e interpunzione del latino nel medioevo. Rome, 1987, 53-68, especially pp. 65-6. Note that there are occasional discrepancies in verse numbering between the Nestle- Aland Greek text (B. Aland, K. Aland et al. (edd.), Novum Testamentum Graece, 27th ed., Stuttgart: Deutsche Bibelgesellschaft, 1993) and the Stuttgart Vulgate (R. Weber, R. Gryson et al. (edd.) Biblia Sacra iuxta Vulgatam versionem, 5th ed., Stuttgart: Deutsche Bibelgesellschaft, 2007); in this survey, the Nestle-Aland num- bering is followed. 2. Two recent studies have shown how the kephalaia are structured around the accounts of Jesus miracles: Greg Goswell, « Early Readers of the Gospels: The Ke- phalaia and Titloi of Codex Alexandrinus » Journal of Greco-Roman Christianity and Judaism 6 (2009) 134-74; James R. Edwards, « The Hermeneutical Significance of Chapter Divisions in Ancient Gospel Manuscripts » NTS 56 (2010) 413-26. 3. Matching numbered paragraphs are also found in the sixth-century palimpsest Codex Zacynthius (G-A 040), extant only in Luke, and minuscule 579. For more details about the history of Greek systems, see Christian-Bernard Amphoux, « La division du texte grec des Évangiles dans l’Antiquité » in Jean-Claude Fredouille et al. (edd.), Titres et articulations du texte dans les œuvres antiques. Paris: Institut des études augustiniennes, 1997, 301-12.
BBene_Boek.indbene_Boek.indb 316316 224/11/20114/11/2011 11:33:2311:33:23 H. A. G. HOUGHTON 317
bian apparatus, a system of concordance between the four Gospels: each Gospel is divided into sequentially-numbered Ammonian sections (John has 232), which are assigned to one of ten Eusebian canons according to the relationship of the material with the other Gospels.4 This appears in Greek biblical manuscripts from the fourth-century Codex Sinaiticus (G-A 01) onwards. The Latin tradition is much more diverse. In Sommaires, divisions et rubriques de la Bible latine (SDR), a collection of material prepared by Donatien De Bruyne in 1914 for the use of scholars working on the Roman edition of the Vulgate, no fewer than fourteen types of chapter divisions are given for John along with an edition of the accompanying titles, or capitula.5 This builds on the information provided in Appendix II of Berger, Histoire de la Vulgate, and entirely replaces the seven series printed by Wordsworth and White in their Oxford Vulgate.6 Nonethe- less, SDR was intended only as an aid for transcribers, lacking introduc- tion, explanation of the sigla, and any analysis of the relationships of the different series: De Bruyne himself later acknowledged that « Le texte était provisoire, le sigle donné à chaque sommaire était souvent arbi- traire. Tout cela était un commencement d’étude, non une conclusion. »7 In the Old Testament it has been superseded as intended by the printed volumes of the Roman Vulgate. For the New Testament, however, the fullest critical account of later Latin tradition is still provided by the Oxford Vulgate and it is only in the Vetus Latina editions that further
4. The Eusebian apparatus and the eighteen-chapter series are printed in the inside margin of Nestle-Aland. Eusebius’ own account of his system is provided in his Letter to Carpianus (in Nestle-Aland, p. 84*). On the significance of the Latin tradition for their transmission, see Walter Thiele, « Beobachtungen zu den eusebi- anischen Sektionen und Kanones der Evangelien » Zeitschrift für die neutestamentliche Wissenschaft 72 (1981) 100-11. 5. [D. De Bruyne], Sommaires, divisions et rubriques de la Bible latine, Namur: Godenne, 1914. On capitula more generally, see Pierre Petitmengin, « Capitula païens et chrétiens » in Fredouille, Titres et articulations du texte, pp. 491-509. 6. S. Berger, Histoire de la Vulgate pendant les premiers siècles du Moyen Âge. Paris: Hachette, 1893 (repr. New York: Franklin n.d.); J. Wordsworth and H.J. White (edd.), Novum Testamentum Domini Nostri Secundum Editionem Sancti Hieronymi. Pars Prior - Quattuor Evangelia. Oxford: Clarendon Press, 1889-98: the first six series of chapter titles (corresponding to SDR Types C, B=A, Br, I, Ifor and Pi respectively) are printed on pp. 492-506; the seventh (corresponding to SDR Type Cat) is in an Appendix on pp. 703-4. 7. D. De Bruyne, « Cassiodore et l’Amiatinus » Revue bénédictine 39 (1927) 261-6 (quotation from p. 264); see also P.-M. Bogaert, « Les particularités éditoriales des Bibles comme exégèse implicite ou proposée. Les sommaires ou capitula donatistes » in Lectures bibliques. Colloque du 11 nov. 1980. Bruxelles: Publications de l’Institutum Judaicum (1982) 7-21.
BBene_Boek.indbene_Boek.indb 317317 224/11/20114/11/2011 11:33:2411:33:24 318 REVUE BÉNÉDICTINE
analysis of this material and improved texts of the capitula are to be found.8 The present article examines the chapter divisions and capitula in the Gospel according to John with reference to the Old Latin tradition. This consists of early Latin versions which differ from the revision by Jerome later adopted as the Vulgate. Old Latin manuscripts of John have been divided by Burton into three groups: Group 1 provides the most ancient texts; Group 2A represents a later, more consistent form; Group 2B com- prises manuscripts closely related to the Vulgate but preserving a sub- stantial proportion of divergent readings.9 The first section considers the systems of division found in these codices, demonstrating that two of the principal series of chapters occur in Old Latin witnesses not included in SDR, while there is further evidence for a third which is apparently unique to the early versions. The second section investigates each of the types of capitula in turn, clarifying details of their attestation and inter- relation and showing how the form of biblical text in several series not only confirms an Old Latin origin but also provides important evidence for early translations of the Gospel.
1. Divisions of the Gospel Text in Old Latin Manuscripts
Certain features of the layout of the Gospels are characteristic of the Vulgate. In his dedicatory letter to Pope Damasus, which precedes the biblical text in many Latin gospel books, Jerome states that his revision has the Gospels in the order Matthew, Mark, Luke and John, following the Greek sequence.10 This enabled him to add the full Eusebian appa-
8. e.g. R. Gryson (ed.), Apocalypsis Iohannis. (Vetus Latina 26/2) Freiburg: Her- der, 2000-3, pp. 62-77, which reduces De Bruyne’s eleven columns to three principal series. 9. See P.H. Burton, The Old Latin Gospels. A Study of their Texts and Lan- guage, Oxford: OUP, 2000, pp. 62-74. The members and nomenclature of the groups have been developed during subsequent work towards a new edition of the Vetus Latina Iohannes. Manuscripts classified as « Old Latin » are those listed in the official Vetus Latina register: Roger Gryson, Altlateinische Handschriften/Manuscrits Vieux Latins. Première partie: Mss 1-275 (Vetus Latina 1/2A) Freiburg: Herder, 1999, with the subsequent addition of VL 9A and VL 11A. 10. The letter is known as the Epistula ad Damasum or Praefatio in Euangelio and sometimes identified by its opening words, Nouum opus: a critical text is printed in the Stuttgart Vulgate, pp. 1515-6. There is a wide variety of other orderings of the Gospels in extant codices and canonical lists. The most common in Old Latin witnes- ses is Matthew, John, Luke, Mark (as in VL 2, 3, 4, 5, 8, 10, 13, 14, 16 and proba- bly also 17): for a full survey, see P.-M. Bogaert, « Ordres anciens des évangiles et tétraévangile en un seul codex » Revue théologique de Louvain 30 (1999) 297-314.
BBene_Boek.indbene_Boek.indb 318318 224/11/20114/11/2011 11:33:2411:33:24 H. A. G. HOUGHTON 319
ratus: the Ammonian sections in the margin of the Gospels were written in black with the relevant canon number underneath in red. Jerome’s explanation of how to refer to the canon tables in the front of the gospel book in order to identify parallel passages suggests that the practice of including the corresponding Ammonian sections from other Gospels in the margin was a later development.11 The list of lections found at the back of many Vulgate gospel books, the Capitulare euangeliorum de cir- culo anni, uses the Ammonian section numbers to identify the readings.12 In a handful of manuscripts, the opening text of each Ammonian section is provided in the initial canon tables.13 A few witnesses to the Type I capitula have Ammonian section numbers in place of chapter numbers in the list of titles.14 Jerome does not mention other types of biblical chapter division in his preface. Nonetheless, synoptic tables of such systems are provided in the second part of SDR based on a large number of Vulgate manuscripts, along with a few witnesses to the earlier versions. There are fourteen separate columns for John (pp. 521-6), although these mask a consid- erable degree of overlap. Types I, Pi, In and Cat are broadly similar, with thirty-six sections (the last of which is absent from Type Pi). Type B, with fourteen divisions, is the most widely-attested system. Type C (45 chapters) is found in Codex Amiatinus and several other well-known Vulgate witnesses, while the remaining five series only appear in one or two manuscripts: Type P (21 chapters), Type Ifor (39 chapters), Type
11. One of the earliest examples of this appears to be VL 23, a tiny fragment from Egypt palaeographically dated to the fifth century, which has next to John 7:28 the Johannine Ammonian section LXXVI, the canon III (apparently written twice), and underneath these the numbers CXII (the corresponding Ammonian section in Matthew) and possibly also the beginning of CLXXIX (the corresponding section in Luke). 12. See also Vladislav Popovic ´ , « Les Évangiles de Split » Bulletin de la société nationale des antiquaires de France 1987 (1989) 266-89, especially p. 278. The Capitu- lare euangeliorum is widely attested from the ninth century onwards. 13. Six are mentioned by Patrick McGurk, Latin Gospel Books from A.D. 400 to A.D. 800. (Les Publications de Scriptorium 5.) Paris-Brussels: Érasme, 1961, p. 85: Codex Brixianus (VL 10); Poitiers, BM 17 (VL 39); Vendôme, BM 2 (VL 40); Troyes, BM 138; Paris, Mazarine 1 (not the current shelfmark); Trier, Seminary 40. To these may be added Laon, BM 437 bis (VL 46), which is very close to the text found in the tables of Codex Brixianus and preserves Old Latin readings. For the Old Latin element in VL 39 and 40, see P. Minard, « Témoins inédits de la vieille version latine des Évangiles. Les canons à initia des évangéliaires de Sainte-Croix de Poitiers et de la Trinité de Vendôme » Revue bénédictine 56 (1945/6) 58-92. Troyes 138 is a different series, heavily redacted and featuring characteristic Vulgate rea- dings. 14. In John, this occurs in Oxford, Bodl. 155; London, BL, Add. 9381; Paris, BnF, lat. 260; Chartres 31; Bern, Burgerbibl. 85 (see Berger, Histoire de la Vul- gate, p. 354). Several of these manuscripts have Insular connections.
BBene_Boek.indbene_Boek.indb 319319 224/11/20114/11/2011 11:33:2411:33:24 320 REVUE BÉNÉDICTINE
Cy (68 chapters), Type Ben (49 chapters) and Type Q (27 unnumbered sections).15 Further details of the sources for each type are given in the second part of this article: the revised table of divisions in the Appendix includes information for the three columns which are blank throughout (Types Br, D and Z) as well as two sets of capitula which are not included (Types W and Vich).16 With the exception of Type Br (which is identified below as Type A) and the addition of Type Win, this study will continue to use De Bruyne’s sigla for each series. The earliest surviving Latin manuscripts of John have neither Euse- bian apparatus nor chapter numbers. Codex Palatinus (VL 2) and Codex Vercellensis (VL 3) indicate sense units by ekthesis, a line projecting into the left margin, at the rate of roughly one for every two modern verses.17 The original portion of Codex Bezae (VL 5), the fifth-century Greek-Latin bilingual written in sense lines, also features ekthesis but less frequently: the total for the whole gospel would be slightly fewer than the Ammo- nian sections, with which there is no correspondence.18 However, Codex Veronensis (VL 4), despite having the Gospels in the Old Latin order and being classified as a Group 1 witness in the first half of John, includes both the Ammonian sections and Eusebian canon numbers in silver and gold ink respectively. While it is possible that the numbers were taken from a different source to the exemplar used for the gospel, their presence in this fifth-century Old Latin codex suggests that Eusebius’ system may
15. There are no corresponding series of capitula for Types P and Q; in Mark to John, the sole witness to the Type P divisions, Paris, BnF, lat. 10348, has the Type B capitula (Sommaires, divisions et rubriques, p. 414). 16. The blank column headed Type X presumably refers to Type Z, as Type X is not present in John. 17. The exact figures are 451 projecting lines in Codex Palatinus and 363 in Codex Vercellensis according to the transcriptions in the Vetus Latina Iohannes Electronic Edition (P.H. Burton, J. Balserak, H.A.G. Houghton, D.C. Parker (edd.) Vetus Latina Iohannes. The Verbum Project. The Old Latin Manuscripts of John’s Gospel. Version 1.6 (January 2010), online at
BBene_Boek.indbene_Boek.indb 320320 224/11/20114/11/2011 11:33:2411:33:24 H. A. G. HOUGHTON 321
also have been adopted in the Latin tradition independently of Jerome. There are 504 instances of ekthesis and 637 gold paragraphus markers in John in Codex Veronensis, but no evidence of division into chapters. The three other members of Group 1, all copied in the sixth or sev- enth century, do not have Ammonian sections but do contain systems of chapters. The fragmentary leaves of Codex Sarzanensis (VL 22) reveal two series of numbered sections. First, there are four Greek numerals, Γ at 3:22, Δ at 4:1, E at 6:1 and Ζ at 7:1, which correspond broadly to the fourteen-chapter Types A and B.19 In addition, there is an over- lapping series of Roman numerals: VI occurs alongside 6:15, and XI to XVIII between 7:2 and 9:1.20 This sequence seems to be without paral- lel, and cannot have extended back to the beginning of the gospel with the same frequency: it is a shame the manuscript is not better preserved to shed more light on this. Codex Monacensis (VL 13) indicates chapter divisions by three rubricated lines plus a decorated capital in the mar- gin, but no numbers. This is the sole witness for Type Q in SDR. There are twenty-seven such divisions in the extant part of the manuscript.21 Finally, Codex Usserianus primus (VL 14) also rubricates the beginning of each section. There are thirty-three of these lines remaining, but num- bers occasionally visible to the left of the rubric indicate that this was a thirty-six chapter sequence very similar to Type I: twenty-six correspond exactly, and six of the seven variants are by no more than a verse. The exception is the placing of Chapter 16: although this occurs in Type I at John 8:1, the story of the adulteress is missing from this manuscript and the rubric is found at 8:12 instead. Nonetheless, as the correspond- ing title in Type I and related series refers only to this pericope, it seems more likely that Codex Usserianus is an accommodation to an existing scheme rather than an earlier version of this set of divisions.
19. Type B is an abbreviation of Type A, and the archetypical divisions appear to be identical (although their location fluctuates in the manuscripts). In the table in SDR, Type B has the fifth division at 5:1 (not extant in VL 22) and the sixth divi- sion at 6:3; it is possible that there was confusion between the Greek numerals E (5) and Ϛ (6). Alternatively, could it be that the division at 6:15 with Roman numeral VI actually represents Ϛ and the division at 5:1 was omitted? 20. Godu, the more recent editor of the manuscript, transcribes the numbers from 16 onwards as XG, XGI etc. using the older G ligature for VI (cf. B. Bischoff, Latin Palaeography [trans. Ó Cróinín and Ganz], Cambridge: CUP, 1990, p. 176). 21. There are also four occasions when a larger capital is found without rubrica- tion (John 1:41, 2:12, 4:51, 6:11), but as these are not decorated, they do not seem to be part of the scheme of division and are not taken as such by De Bruyne. Several inaccuracies in the table in SDR have been corrected in the Appendix below.
RB 22
BBene_Boek.indbene_Boek.indb 321321 224/11/20114/11/2011 11:33:2411:33:24 322 REVUE BÉNÉDICTINE
Three of the four principal members of Group 2A also have the chap- ter divisions of the Type I group: Codex Colbertinus (VL 6), Codex Cor- beiensis (VL 8) and the mixed-text Old Latin manuscript in Würzburg (VL 11A).22 All six marginal numbers in the last of these occur between John 3:1 and 5:1, one of the portions taken from an Old Latin exemplar. Neither VL 11A nor VL 8 have Ammonian sections, but the much later VL 6 has the full Eusebian apparatus including Ammonian sections for the parallel passages in the Synoptic Gospels. There are minor differences between the location of the divisions in VL 6, VL 8 and the archetypical version of Type I in SDR; some of these match other sequences, but the capitula indicate that they are witnesses to Type I.23 The fourth manu- script in Group 2A, Codex Rehdigeranus (VL 11), has the Ammonian sec- tions for John alongside the gospel text and those of the parallel passages in a decorative arcade at the bottom of the page (without any Eusebian canon numbers). Although it has the Gospels in the Vulgate order, this presumably represents a method of applying the Eusebian apparatus to the Latin tradition without the canon tables. There are also six marginal numbers representing longer divisions: 2 (at 2:1); 3 (at 3:22); 5 (at 4:54); 6 (at 6:1); 9 (at 7:1) and 11 (at 13:1).24 Four of these correspond to Types A and B (2, 3, 6 and 11); 5 is only one verse away from the expected loca- tion at 5:1, and 9 (VIIII) seems to be an error for 7 (VII), in keeping with comparable mistakes in the Ammonian sections.25 This manuscript may therefore be taken as a second Old Latin example of Type A or B, complementing the Greek numerals in VL 22.26 The only one of the fragmentary members of Group 2A with sectional divisions is the St Gall fifth-century uncial (VL 16). This does not have the Eusebian apparatus, but Greek chapter numbers have been added in red. The single complete page of John 19 has IOHANNEN ΙΔ as the running title, the fourteenth chapter matching the Type A and B divisions; the more extensive numer-
22. For this, see H.A.G. Houghton, « A Newly Identified Old Latin Gospel Manus- cript: Würzburg Universitätsbibliothek M.p.th.f. 67 » JTS ns 60.1 (2009) 1-21. 23. The full list of divisions is supplied in the Appendix; both manuscripts are lacking Chapter 36. 24. Some appear to be combined with rubricated lines, although it was not always possible to identify these with confidence from the monochrome microfilm used for this study; there do not appear to be rubrics accompanying chapters 5 and 9, while there may be rubricated lines elsewhere not related to the numbered divisions. 25. For example, XCII and XCIIII for XVII and XVIIII at 1:41 and 2:12; LI for XLI at 5:24; LXXVII erroneously at 8:14 (between LXXXVI and LXXXVII); CXXXVIII for CXXXVII at 15:16. 26. Type B is supported by its affiliation in the Synoptic Gospels: see further the section below on Type B capitula.
BBene_Boek.indbene_Boek.indb 322322 224/11/20114/11/2011 11:33:2411:33:24 H. A. G. HOUGHTON 323
ation of Matthew and Mark confirms this identification.27 It is interesting that, as in VL 22, Greek numerals (and the Greek accusative) are used for this series.28 Manuscripts in Group 2B, being close to the Vulgate, usually include the full Eusebian apparatus and the Gospels in their customary sequence. Codex Brixianus (VL 10) is an exception on both counts, having the order Matthew, John, Luke, Mark and only the Ammonian sections in the mar- gin, with the Synoptic parallels indicated in arcades at the foot of each page (like VL 11). The canon tables at the beginning of the manuscript include the opening words of each Ammonian section (and occasional subdivisions): these appear both codicologically and textually to derive from a different source, preserving some Old Latin readings not matched by the gospel text although their character is broadly similar.29 Codex Brixianus features ekthesis in 111 lines in John, as well as a series of lon- ger numbered divisions in the first half of the Gospel which are preceded by the letters LEC. Nine of the first ten divisions match the rubrics in Codex Monacensis which constitute Type Q: LEC I appears at John 1:35, corresponding to the second rubric (the first is at John 1:1); the numbers are made up, however, by LEC III at John 2:12 where there is no rubric in Codex Monacensis.30 The systems diverge from the middle of John 6: LEC XII at 6:47, LEC XIIII at 7:14, and LEC XVIII at 9:39 have no counterparts in Codex Monacensis, while LEC XVI at 8:20 anticipates the rubric by one verse. From John 10, there are no LEC markings until a unnumbered LEC added by a later hand at 20:19. Even so, the partial similarity of these systems is significant, suggesting that Type Q may not have been peculiar to Codex Monacensis. It is worth noting in pas- sing that both these manuscripts have Arian connections, although they
27. In the fragments of Matthew ΙΗ (18) appears next to 18:1 and in the running title of the next page, Κ (20) next to 20:1 and ΚΑ (21) alongside 21:1, while in Mark Ϛ (6) can be detected in the running title of the page beginning 7:13 and possibly also Ζ (7) on the page beginning 8:32 (see St Gall 1394, pp. 52, 66, 62, 58, 75 and 79). The size of numbers in the running titles suggests that they were added after the copying of the gospel text, although it is hard to say how much later: the numbers in Mark have faded and it is impossible to detect one in the running title of Mark on p. 85 or its seventh-century replacement page (p. 91). 28. The use of Greek numerals for the capitula in certain witnesses of Cyprian’s collection of testimonia and other Latin manuscripts is noted in Parker, Codex Bezae, p. 9. Colmar 38 also identifies the chapter divisions with Greek numerals, preceded by K: see Vezin, « Les divisions du texte », pp. 57-8. 29. See Roger Gryson, « La version gotique des évangiles. Essai de réévaluation » Revue théologique de Louvain 21 (1990) 3-31, p. 23. The text of these titles is included with that of the other Old Latin canon tables in the Vetus Latina Iohannes. 30. As noted above, there is a larger capital at this point in VL 13, although this is one of several which do not seem to relate to the rubricated divisions.
BBene_Boek.indbene_Boek.indb 323323 224/11/20114/11/2011 11:33:2411:33:24 324 REVUE BÉNÉDICTINE
belong to different textual groups.31 The designation LEC is presumably an abbreviation for lectio, reminiscent of manuscripts which divide the text of the gospels into paragraphs with the liturgical incipit in illo tem- pore: the only Old Latin manuscript with this is Codex Sangermanensis secundus (VL 29). This has fifty-three lections marked within the text of John, sometimes inserting introductory phrases into the middle of a verse regardless of the context.32 Three other members of Group 2B do not have any of the Eusebian apparatus. The portions of John surviving in Codex Usserianus secundus (VL 28) feature frequent capital letters but no marginal numbers. St Gall 51 (VL 48) is similar, with over six hundred rubricated capital letters. Some extend over two or more lines, with exceptionally large letters or groups of letters at 1:1, 1:6, 10:1, 13:1, 18:1 and 20:1. McGurk identifies several manuscripts of insular origin which share this feature.33 It is dif- ficult to specify which (if any) of the other larger capitals are particularly significant: sixteen more occur at the beginning of a line, displacing text below, but this is occasionally the case in the middle of a line as well (e.g. 7:1, 21:1) and there are also larger capitals which do not affect the following line (e.g. 2:1, 2:12, 3:22). While most are paralleled by sections found elsewhere, there is no obvious correspondence with any single sys- tem of division given in SDR.34 The only form of division in the tiny sixth-century copy of John found in a reliquary in Chartres (VL 33) is ekthesis, which occurs in 308 lines. Divisions corresponding to Type I and Types A and B are also pre- sent in Group 2B. The former is found in Codex Sangermanensis pri- mus (VL 7, but practically a Vulgate text in John) and Codex Aureus
31. For VL 13, see Roger Gryson, « Les citations scripturaires des œuvres attri- buées à l’évêque arien Maximinus » Revue bénédictine 88 (1978) 45-80; for VL 10, which is believed to originate from a Latin-Gothic bilingual, see F.C. Burkitt, « The Vulgate Gospels and the Codex Brixianus » JTS 1 (1900) 129-34 and Gryson, « La version gotique des évangiles ». 32. Although VL 29 has the Type Pi titles for John, the corresponding divisions do not appear in the text of the Gospel. A particularly good example of the liturgical arrangement is Florence, BML, Plutei 25.2, which has 39 unnumbered paragraphs in the extant part of John: all except three are paralleled in VL 29, but there is no match with any of the series in SDR. Note too the use of capitula lectionum for the chapter titles in Type C (see below). 33. McGurk, Latin Gospel Books, pp. 117-9. Two of these manuscripts also have an outsize capital at John 2:1. 34. Thirty double-height (or larger) capitals are indicated in the table in the Appendix, although this selection is necessarily arbitrary.
BBene_Boek.indbene_Boek.indb 324324 224/11/20114/11/2011 11:33:2411:33:24 H. A. G. HOUGHTON 325
(VL 15). 35 Again, there are several slight variations, usually by no more than a verse. St Gall 60 (VL 47) only contains John but has the full Eusebian apparatus including Synoptic parallels as well as fourteen num- bered chapters. The latter are an almost exact match for the divisions of Types A and B.36 In addition, there is a single fossilised number cor- responding to the Type I series copied as part of the biblical text in the middle of John 2:13, showing that this series was present in an ances- tor.37 The newly-identified Old Latin manuscript in St Petersburg, VL 9A, also has fourteen rubricated lines and numbered divisions in the text of John (notwithstanding the initial list of forty-five capitula).38 The first eleven correspond more or less to Types A and B, albeit increased by one because the first chapter is numbered 2. There is no division at 13:1 or 15:1; instead Chapter 12 comes at 16:33, Chapter 13 at 18:1 and Chapter 14 in the normal place at 18:28. Finally, St Gall 48 (VL 27) is not a self-standing Latin witness but a Greek gospel book with an interlinear Latin translation. This has the standard series of Greek kephalaia: the eighteen titles occur both in an initial list and within the gospel text, and the Latin equivalent is pro- vided above. In conclusion, despite the absence of systematic chapter divisions from several of the oldest Latin gospel books, the two best-attested systems of dividing John, a thirty-six chapter sequence (Type I, cf. Pi, In, Cat, W) and a fourteen-chapter sequence (Type A or B, cf. Vich), are both found in Old Latin manuscripts with the earliest form of text (Burton’s Group 1). This is a significant advance on the evidence given in SDR for tracing the development of these divisions. Similarly Type Q, which was previously known only from Codex Monacensis (VL 13), also a member of Group 1, finds further support in the numbered lections of the first half of John in Codex Brixianus (VL 10). Although the other types of division listed in SDR do not appear to be present in manuscripts identi-
35. These are presented in the Appendix. There is a group of manuscripts closely related to Codex Aureus in John which also have Type I capitula and divisions (see Type I and Type In below). 36. Note that this is based on the divisions as given in the present Appendix rather than SDR, which is erroneous at a number of points in Type B. VL 47 also has 358 double-height rubricated capitals at the beginning of sentences, some of which correspond to Ammonian sections. 37. The only Type I witness beginning Chapter 5 with et ascendit is VL 14, another Insular manuscript; see the Appendix. 38. See H.A.G. Houghton, « The St Petersburg Insular Gospels: Another Old Latin Witness » JTS ns 61.1 (2010) 110-27 (especially p. 114) and the section on Type C below.
BBene_Boek.indbene_Boek.indb 325325 224/11/20114/11/2011 11:33:2511:33:25 326 REVUE BÉNÉDICTINE
fied as Old Latin, the larger capitals in St Gall 51 (VL 48) indicate that yet more configurations may be found. None of the known Latin sys- tems, however, correspond to the Greek kephalaia or the paragraphs in Codex Vaticanus.39 Furthermore, no two witnesses have exactly the same distribution of divisions: this suggests that exact agreement even within the same overall scheme is the exception rather than the rule and could therefore be of genealogical significance.40 The presence of chapter num- bers, especially incomplete series, may provide an indication of the use of different exemplars (as in the case of VL 11A), or shed other light on the transmission of the text. These systems might also convey information on how the text of the Gospel was approached by exegetes or the way in which gospel books were used.41 While the existence of chapter divisions in early Latin biblical manuscripts has long been known, the extent and variety of their attestation in witnesses with an Old Latin text of John is remarkable.
2. The text of the capitula lists
Accompanying each set of chapter divisions is a list of chapter titles, or capitula. (The individual titles may be referred to as tituli.) These were part of Latin biblical tradition from at least the middle of the fourth century: Jerome mentions a series of titles for the Gospels attributed to Fortunatianus of Aquileia, and Hilary of Poitiers’ commentary on Mat- thew includes capitula very similar to those preserved in some Gospel manuscripts.42 The capitula lists usually precede the text of each Gospel,
39. Berger’s attempts to connect Type I or Type A/B in Latin Gospels with the Vaticanus paragraphs are misleading (Histoire de la Vulgate, pp. 311-2): similarities are scarcely surprising given the narrative shape of the text, but the high propor- tion of Vaticanus divisions without a Latin counterpart tells against their use as a source. 40. Variations in Ammonian section numbers are similar indicators: see Patrick McGurk, « The Disposition of Numbers in Latin Eusebian Canon Tables » in R. Gry- son (ed.), Philologia Sacra I. Altes und Neues Testament. Freiburg: Herder, 1993, pp. 242-58 (reprinted in P. McGurk, Gospel Books and Early Latin Manuscripts Aldershot: Ashgate, 1998), and Vladislav Popovic ´ , « Du nouveau sur les Évangiles de Split » Bulletin de la société nationale des antiquaires de France 1990 (1992) 275-93, pp. 290-1. 41. For patristic use of the term capitulum, see Petitmengin, « Capitula païens et chrétiens »; for Augustine in particular, H.A.G. Houghton, Augustine’s Text of John. Patristic Citations and Latin Gospel Manuscripts. Oxford: OUP, 2008, pp. 40-1. 42. Fortunatianus, natione afer, Aquileiensis episcopus, imperante Constantio in euangelia titulis ordinatis breui sermone et rustico scripsit commentarios. (Jerome, De uiris illustribus 97, ed. E. Richardson, TU XIV.Ia, Leipzig, 1896); see further Berger, Histoire de la Vulgate, pp. 308-9.
BBene_Boek.indbene_Boek.indb 326326 224/11/20114/11/2011 11:33:2511:33:25 H. A. G. HOUGHTON 327
although in some codices they occur in a separate section preceding all four (e.g. VL 7). While the presence of divisions in a gospel manuscript does not offer a secure indication of its chronology or origin, the match- ing capitula have distinctive textual characteristics, particularly in their quotation of biblical material, which may make it possible to propose a location or date for their original composition. In this section, each of the fourteen sets of capitula printed in SDR (along with a fifteenth men- tioned by Berger) is analysed to see whether it is a fresh compilation or dependent on another series, and whether its biblical text derives from an Old Latin or Vulgate source. In addition, full details are given of the manuscripts used by De Bruyne for his edition (which are not always easy to identify), supplemented by information from McGurk’s catalogue of Latin gospel books copied before the year 800.43 The series are treated in alphabetical order; only the opening and closing capitula are quoted here as a full text of the Old Latin types is now available in the first fascicule of the edition of the Vetus Latina Iohannes.
Types A and B In the Synoptic Gospels, separate series of capitula are given as Type A and Type B, although they are very closely related: Type B is an abbre- viation of Type A.44 In John, however, SDR presents a single set with the siglum B=A, on the grounds that many manuscripts with Type A in the Synoptics have the shorter titles typical of Type B in John. In fact, the textual relationship between De Bruyne’s Type B=A and Type Br is comparable to that of B and A in the Synoptic Gospels; furthermore, this had already been recognised by Berger, who treated the Br titles as the original form of Type A although the shorter titles for John had clearly been substituted at an early point in at least one branch of the tradi- tion.45 In fact, the summary tables in SDR (pp. 411-4) preserve the Type A and Type B distinction found in Berger, apparently ignoring Berger’s note that, while at least twenty-four of the otherwise pure Type A wit-
43. The capitula of John are presented in the following order in SDR: Types D, I, W, Cat, Ifor, Pi on pp. 264-9; Types B=A, In, Br, Cy, C, Z, Vich, Ben on pp. 302-11; in addition, the tables of affiliation on pp. 411-4 are often useful. McGurk, Latin Gospel Books, pp. 113-17 includes nine of the fourteen series of capitula. 44. Popovic ´ , « Du nouveau sur les Évangiles de Split », p. 285, states that A stands for Antiquus and B for Breuiatus and that they are probably of Roman origin. 45. See the table in Berger, Histoire de la Vulgate, p. 356; the introduction of the shorter version for John may derive from a single Vulgate manuscript which used a different exemplar for this Gospel but proved to be influential in the subsequent transmission of the text.
BBene_Boek.indbene_Boek.indb 327327 224/11/20114/11/2011 11:33:2511:33:25 328 REVUE BÉNÉDICTINE
nesses have Type B in John, there are around forty manuscripts in which the affiliation of John requires verification (see below). In the present study, they will once again be divided into the abbreviated Type B and the longer Type A (formerly Type Br), taken in this order.
Type B (formerly Type B =A). 14 chapters. Incipit (I): Pharisaeorum leuitae interrogant Iohannem, Iohannes Iesum uidens agnum dei dicit et Andreas Petro dicit: inuenimus messiam. Explicit (XIIII): Adlocutio Pilati ad Iudaeos de Iesu et de Barabba. Passio Iesu et sepultura et resurrectio eius.46
De Bruyne edits the text of the Type B capitula from sixteen manu- scripts.47 McGurk supplements these with sixteen instances of this series in gospel books copied before the year 800, the earliest of which is the sixth-century Gospels of St Augustine.48 These titles occur in two manu- scripts with more than one series of capitula: VIENNA, lat. 1224, where they immediately precede the Gospel (after Type D at the beginning of the manuscript), and Codex Forojuliensis, where they are followed by Type Ifor. The sole example of Type P divisions (Paris, BnF, lat. 10348) has Type B capitula. The titles are preceded by a variety of headings: the majority of manuscripts describe them as capitula, but we also find breues, breuiarium, elenchus and tituli.49
46. Like Type A, many manuscripts with Type B capitula begin the fourteenth title with Passio, and have adlocutio ... Barabba as the final sentence of the thir- teenth title. The placing of the fourteenth division alongside 18:28 in the gospel text, however, supports De Bruyne’s disposition of the titles. 47. These are Codex Cavensis (C; Cava, Badia 1); Toletanus (T; Madrid, BN, Vitr. 13-1/Tol. 2.1); Complutensis 2 (t; Madrid, Bibl. Univ. 32); Autun 3 (A); Paris, BnF, lat. 256 (D); Paris, BnF, lat. 17226 (N); Ingolstadiensis (I; Munich, Univ. 2o 29); Forojuliensis (J; Cividale, Mus. arch. etc.); Oxford, Bodl., Auct. D.ii.14 (O); Cambridge, Corpus Christi 286 (X; Gospels of St Augustine); Theodulfianus (Θ; Paris, BnF, lat. 9380); Vatican, Vat. lat. 5645 (R?); Vatican, Vat. lat. 43 (V); Vatican, Palat. lat. 46 (P); Ivrea, Bibl. cap. 99 (Y); Turin F.vi.I (Z). The manus- cript with siglum R is not identified on p. 270, although a comparison with its text in the final capitulum as reported by McGurk and a handwritten note in De Bruyne̕ s own copy observed by P.-M. Bogaert indicate that it is Vatican lat. 5645. Note also that Vatican Pal. 46 and Pal. 48 seem to have been swapped on pp. 411-2. 48. Seven of these are used by De Bruyne: witnesses A, D, N, J, O, X, R in the preceding note. The others are Abbeville 4; Autun 4; Cambridge, Univ., Kk.I.24; Colmar 38; London, BL, Add. 5463 (mistyped as 5436 on p. 117); London, BL, Harley 2788; Paris, Arsenal 599; Trier, Stadtbib. 22; Vienna, lat. 1224 (second series). 49. Breues: J (first series), Θ, A (but expliciunt capitola), D, Autun 4; breuiarium: Harley 2788, Vatican lat. 5465; elenchus C, T, N, Vienna 1224 (second series); tituli: O has no heading, but ends expliciunt tituli. On the introduction of the term breuiarium by Sedulius Scottus, see Petitmengin, « Capitula païens et chrétiens »,
BBene_Boek.indbene_Boek.indb 328328 224/11/20114/11/2011 11:33:2511:33:25 H. A. G. HOUGHTON 329
As noted above, VL 16, VL 22 and COLMAR 38 use Greek numerals to mark chapter divisions matching this series, although the Old Latin wit- nesses do not transmit the capitula. VL 11, however, has Type B capitula before Luke and it is possible that Type B titles originally preceded John in the pages now missing.50 An Old Latin origin for the series is sup- ported by the biblical text in nine of the fourteen titles, e.g. minorari in Cap. 3 (VL 4, 14; Vulgate minui) and both dedicatio (VL 2, 6, 8, 10, 14, 32, 46; Vulgate encenia) and deambulare (VL 11, 11A, 14; Vulgate ambu- lare) in Cap. 9. However, all such non-Vulgate readings are shared with the longer capitula of Type A which have an even more marked Old Latin affiliation (see below). The two series are obviously related: much of the wording is identical, including phrases such as de grano frumenti quod in terram mittitur (Cap. 10), de obseruandis mandatis (Cap. 12) and adlocutio Pilati ad Iudaeos de Iesu (Cap. 14). Confirmation of the dependence of B on A may be found in two summary phrases: Cap. 2 in B ends et de aliis multis, abbreviating the mention of the serpent and the light in A, while Cap. 12 in B ends et cetera mandata, which are more fully listed in A.51 The material in Type B which does not derive from Type A is minimal: the most striking is the insertion in Cap. 5 of scrutamini scripturas and si crederetis Mosi crederetis forsitan et mihi (John 5:39, 46). Both these quotations correspond to the Vulgate, as does the replacement of initium by principium in Cap. 8 (cf. John 8:25). There is also the addition of Scenophegia in Cap. 7 and Barabbas in Cap. 14. Although references to the woman taken in adultery and the final chapter of John are missing from the editorial text in SDR, these are unlikely to be textually signifi- cant given that Type B is derived from Type A.52
p. 496; Meyvaert suggests that breues may be an indication of Insular origin (Paul Meyvaert, « Bede’s Capitula Lectionum for the Old and New Testaments » Revue bénédictine 105 (1995) 348-80, p. 350). 50. See McGurk, Latin Gospel Books, pp. 94 and 114, who also notes that it has Type I capitula before Mark. 51. Compare the phrase et ceteris mandatis in Type B in Matthew Cap. 4 and Mark Cap. 12 (and et mandatis in Luke Cap. 5), where Type A provides more details; simi- larly et reliqua at the end of Luke Cap. 20. 52. In the text printed in the Vetus Latina Iohannes, Type B includes de muliere adultera, supported by the majority of manuscripts. For the omission of the last chapter, compare the relationship between Types A and B in Luke. It should also be noted that SDR lacks the phrase et duobus piscibus from Cap. 6 through a typo- graphical error.
BBene_Boek.indbene_Boek.indb 329329 224/11/20114/11/2011 11:33:2511:33:25 330 REVUE BÉNÉDICTINE
Type A (formerly Type Br). 14 chapters. Incipit (I): Pharisaeorum53 leuitae interrogant Iohannem. Iohannes Iesum uidens ecce agnus dei dicit de cuius spiritu testificatur. ... ubi etiam ad Nathanahel loquitur. Explicit (XIIII): Allocutio Pilati ad Iudaeos de Iesu ... et quod post pran- dium Petro pascendas oues tertio iniungit.54 Four manuscripts are listed for Type Br in SDR, yet only Flo - rence, BML, Plutei 25.2 (F) and Vatican, Reginenses lat. 14 (R) are used for the text; Rouen A.1 also contains the full series, but further investigation of Paris, Arsenal 33 reveals that it switches to Type B from the third title of John onwards.55 However, as noted above, Berger lists around forty manuscripts with Type A in the Synoptics which may have this series in John. It was beyond the scope of the present study to investigate all these, but London, British Library, Royal 1.D.III is also a complete example of Type A in John, while Paris, Mazarine 2 is virtu- ally identical to Arsenal 33.56 Five of these six witnesses are two-volume Bibles produced in the tenth century or later (the exception being R). Again, all but one have Type A in the Synoptics, confirming the unity of the series.57 The numeration of the titles varies considerably: in the Florence and Rouen manuscripts the capitula are unnumbered and are only distinguished by (respectively) six and fifteen lines with ekthesis; the numeration of Arsenal 33 is inconsistent (perhaps reflecting its compos- ite nature), but Mazarine 2 numbers the titles in sequence. The Vatican manuscript supports the presentation of Type Br in fourteen sections, although the additional division in SDR with 63 hanging lines, broadly
53. In SDR, the opening text of Type Br is given as Pharisaeorum sacerdotes
BBene_Boek.indbene_Boek.indb 330330 224/11/20114/11/2011 11:33:2511:33:25 H. A. G. HOUGHTON 331
based on the larger capitals in this witness, seems unwarranted. (It is not supplied for Type A in the Synoptic Gospels.) Notwithstanding dif- ferences in De Bruyne’s edition (and further manuscript variation), the number and location of the headings in Types A and B seem originally to have been identical. The earliest description of the Type A titles is breuis digestio euangelii cata Iohannem; the Vatican, Rouen and Arsenal manuscripts have capitula, while Mazarine 2 has breuiarium. The Grae- cism cata serves as an indication of their antiquity. The precedence of Type A over Type B is demonstrated by the marked Old Latin affiliation displayed in its more extensive biblical quotations. Readings characteristic of the earlier versions include in solitudine (3:14; VL 4, 14, 47) and lumen in Cap. 2 (3:19; cf. VL 4, 13, 22, 47), natatoria in Cap. 5 (5:2; VL 2, 3, 4, 5, 8*, 9A*?, 11, 11A, 14), lumen mundi in Cap. 7 (8:12; VL 14), initium in Cap. 8 (8:25; VL 3, 5, 6, 14, 15, 22), accipiendi in Cap. 9 (10:18; VL 11), graeci in Cap. 10 (12:20; VL 2, 3, 4, 5, 6, 8*, 11, 14, 46), uinea (15:1 or 15:4; VL 14), parturiente (cf. 16:21; cf. VL 3, 9A*, 13), tribulatio and saeculo in Cap. 12 (16:33; VL 14, cf. VL 2, 5, 6, 13), ho - norificare in Cap. 13 (17:1-5; VL 3, 5, 13, 14) and retia and nauis in Cap. 14 (21:6; cf. VL 2, 5, 10, 13, 14), in addition to those already discussed for Type B. With such a high proportion of Old Latin forms, it is almost inconceivable that these capitula are a later expansion of Type B rather than the source from which they were abbreviated. The agreements with VL 14 indicate that the Old Latin exemplar had an early form of text (Group 1). For this reason, readings such as inuitatus ad cenam in Cap. 2 (2:2; VL 4, 14, 47) and ascendit in medio die festo in Cap. 7 (7:14; cf. VL 2, 9A and Type B) have a strong claim to be the original text of the series. Adjustment towards the Vulgate is hardly surprising in these compara- tively late witnesses: readings such as frumenti in Cap. 10 (12:24) and dilectione in Cap. 12 (15:13) may have replaced earlier forms. Direct quotation is comparatively rare in these capitula. Instead, a narrative has been created which makes clear reference to individual verses, although the gospel text is sometimes treated rather loosely. For example, dicunt ei quod Iesus omnes baptizaret in Cap. 3 runs together the two parts of John 3:26; prophetam in patria sua sine honore esse in Cap. 4 is closer to Synoptic parallels such as Matthew 13:57 than propheta in patria sua honorem non habet at John 4:44; mittere is found in Chromatius’ references to John 12:24 but biblical codices all have the verb cadere; the sense of Jesus’ teaching at John 15:13 is not completely represented by et dicit nihil maius dilectione (Cap. 12); in John 17 Jesus never uses the verb
BBene_Boek.indbene_Boek.indb 331331 224/11/20114/11/2011 11:33:2511:33:25 332 REVUE BÉNÉDICTINE
δοξάζειν of the disciples, despite the summary ut honorificentur a patre postulat (Cap. 13). This suggests that the compiler was more interested in the general import of passages than an exact representation of the evangelist’s words. It has long been recognised that certain capitula for Acts and the Hebrew Prophets were composed in Donatist circles.58 Although there is nothing distinctively African about the biblical text of this set for John, a number of topics pertinent to the Donatist controversy recur in the relatively limited selection of passages. There is mention of baptism in Cap. 3 and twice in Cap. 4; several references are made to disagreements (Capp. 5 and 7), to treachery and denial (Capp. 7, 11 (twice), 13 (twice, including the term traditio) and 14), and to plots and violence (Capp. 9, 10, 13). Themes such as de tribulatione in saeculo toleranda (Cap. 12) and the Jews’ acclamation of Caesar (John 19:15; Cap. 13) could also be connected with a Donatist context. Alternatively, it has been claimed that one of Types A or B represents the summary associated with For- tunatianus.59 In either case, the evidence is insufficient to reach a firm conclusion.
Type Ben. 50 chapters. Incipit (I): De principio uerbi. Explicit (L): De apparitione eius60 ad mare Tiberiadis et Petro dicit: diligis me. The fifty short titles of Type Ben are only known from four manu- scripts, of which three are used for the edition in SDR: Rome, Casana- tensis 1101, which has the series in all four Gospels; Monte Cassino 35, containing all but Matthew; the sixth- or seventh-century gospel book in Split Cathedral, the oldest witness, with Type I in the Synoptics but Type Ben in John. One further witness not included in SDR is Rome,
58. See Bogaert, « Les particularités éditoriales », pp. 9-10: the first set of capitula for Acts in the Oxford Vulgate (beginning De passione et resurrectione domini) feature references to rebaptism, while Donatist summaries for the prophets are identified in H.J. Chapman, « The Codex Amiatinus and Cassiodorus », Revue bénédictine 39 (1927) 12-32; De Bruyne’s suggestions for other potential Donatist series are mentioned under Type Ben below. 59. Thus Berger, Histoire de la Vulgate, p. 312, who asserts Fortunatianus’ au- thorship of Type B based on the form used by Hilary of Poitiers. Yet it is not clear from Jerome’s text whether the tituli ordinati were composed by Fortunatianus himself (or, indeed, whether they refer to the Gospels or the commentaries); further- more, does the abbreviated form constitute a separate work? 60. Eius is duplicated in SDR.
BBene_Boek.indbene_Boek.indb 332332 224/11/20114/11/2011 11:33:2511:33:25 H. A. G. HOUGHTON 333
Angelica 29, from the tenth century.61 Each title begins with ubi or de, and in the Split codex they have the heading capitula. Although most of the locations of division are similar to other series, Type Ben appears to be a fresh compilation: none of the other thirteen types quotes from John 1:16, 5:41, 6:17, 8:22, 9:39 or 13:37. Interestingly, two of these have Old Latin readings: honorem ab hominibus non quae- rendum in Cap. 13 (cf. 5:41 and 44 in VL 4, 11, 14, 47) and de ascensione nauiculae in Cap. 14 (cf. 6:17 in VL 4, 10, 13, 14). Furthermore, there are at least six other distinctive non-Vulgate forms: triduo in Cap. 6 (2:19), heremo in Cap. 7 (3:14), ante Abraham ego sum in Cap. 24 (8:58), ianua in Cap. 26 (10:1), copiosum in Cap. 33 (12:25) and proicient in Cap. 40 (16:2). Three of these are not present in surviving Old Latin manuscripts: he - remo is found in John 3:14 (and 1:23) in several early authors and is read at 6:39 by VL 14;62 copiosum for πολύν has parallels at 5:6 (VL 11, 14), 15:5 (VL 3, 6, 8, 13) and 15:8 (VL 13) but not 12:25;63 proicient in 16:2 is comparable to expulerunt (VL 2), eicient (VL 5, 11, 13) and expellent (VL 10), but contrasts with facient in the Vulgate and other manuscripts.64 Alongside these should be set forms characteristic of a later stage in the Latin Bible, such as numquid et uos uultis abire (6:67, Cap. 17), de grano frumenti (12:24, Cap. 33) and ubi Iesus crucem baiulat (19:17, Cap. 46). Nonetheless, the Old Latin readings suggest that this series has a pre- Vulgate origin. At one point, De Bruyne considered that the Type Ben gospel capitula might be of Donatist origin, but the evidence adduced is minimal, and Types A and I have stronger claims in John.65
61. See further Popovic ´ , « Les Évangiles de Split » and, especially, « Du nouveau sur les Évangiles de Split », p. 290, which includes Angelica 29 in this group. Accor- ding to Popovicˊ , the siglum Ben stands for Beneventanus and the switch to this type in John represents the more prestigious of the two codices used for the Split Gospels (« Du nouveau », pp. 283-5). 62. See H.A.G. Houghton, Augustine’s Text of John, Oxford: OUP, 2008, p. 209; eremo occurs in 3:14 in two early gospel books, Cambridge, UL, Kk.I.24 (with Type B capitula) and Würzburg, Univ., M.p.th.q. 1a (with Type W capitula). 63. There are no manuscripts with this reading in Bonifatius Fischer, Die latei- nischen Evangelien bis zum 10. Jahrhundert. IV. Varianten zu Johannes (AGLB 18). Freiburg: Herder, 1991. Nonetheless, it is suggested by three patristic references to John 12:24: AU Ps 19.5.6, MAXn s 3 and PS-HI bre 40. (Abbreviations of patristic works are given according to the Vetus Latina system: see Roger Gryson, Répertoire général des auteurs ecclésiastiques latins de l’antiquité et du haut moyen âge. (Vetus Latina 1/1). 2 vols. Freiburg: Herder, 2007). 64. Compare also proicere in a similar context in VL 14 at 9:22. 65. See Bogaert, « Les particularités éditoriales », pp. 11-16. The only item of characteristic African vocabulary identified in John is arguere in Type Ben (Cap. 42), also present in Type C (Capp. 21 and 29). Apart from insinuat in Type Vich (Cap. 13), none of the other words in this list appears in any of the capitula of John.
BBene_Boek.indbene_Boek.indb 333333 224/11/20114/11/2011 11:33:2611:33:26 334 REVUE BÉNÉDICTINE
Type C. 45 chapters. Incipit (I): In principio uerbum deus apud deum per quem facta sunt omnia et Iohannes missus refertur ante eum qui recipientes se facit filios dei per gratiam suam. Explicit (XLV): Usque tertio dicit Petro amas me, quia ter eum negauerat ... quod crucis morte foret martyrio coronandus. De Bruyne lists 24 manuscripts with Type C capitula for John, and uses the same five famous codices as the Oxford Vulgate for the text: Amiatinus (A; Florence, Amiat. 1); Hubertianus (H; London, BL, Add. 24142); Vallicellianus (V; Rome, Vallicell. B.6); Lindisfarne (Y; London, BL, Cotton Nero D.IV); London, BL, Harley 2797 (L). The only other pre ninth-century witnesses are London, BL, Royal 1.B.VII and the St Petersburg Insular Gospels (VL 9A): in the latter, the discrepancy between the Type C capitula and the Type B divisions was caused by the pages of prefatory material being copied separately and added later.66 In VL 9A and Lindisfarne the titles are described as capitula lectionum, a designation which appears to be peculiar to Type C and is found preced- ing the Synoptic Gospels in Amiatinus.67 Other manuscripts have breues (Royal 1.B.VII, although it has capitula lectionum before Luke) or simply capitula (Vallicellianus). The titles and divisions consist of one or two sentences, often heavily subordinated: ablative absolutes and relative clauses are common, while gerunds and gerundives are a recurring stylistic feature.68 The use of inter multa (Capp. 6, 16, 23, 27, 32, cf. 25) and plurima (Capp. 8, 16, 19, 24) indicates that the aim was to summarise the contents of each section. The verbs pronuntiare (Capp. 11, 17, 26, 31), confirmare (Capp. 7, 21,
Despite accepting an African origin for this set, Popovic ´ (« Du nouveau », p. 290) also observes features typical of Rome and Campania. 66. See further Houghton, « The St Petersburg Insular Gospels », p. 114. The other seventeen manuscripts are: Douai 16; Florence, Laurent. 26.1 (given as 25.1 on p. 599); Liège, 1; London, BL, Add. 17738; London, BL, Harley 2788 and 2804; Namur, Seminary; Paris, BnF, lat. 111, 265, 271 and 15177; Metz 1151; Reims 2; Rome, Vallicell. A.2 and D 8; Rome, St Peter D 153; Rouen 2. Although SDR lists Utrecht (Univ. 32) as a witness to Type C in all four Gospels, McGurk, Latin Gospel Books, pp. 77 and 114, confirms that the capitula only survive for Matthew. 67. The heading of the titles for John has been erased in Amiatinus and Hubertia- nus; the titles for Luke in Amiatinus are introduced as indicia siue capitula lectionum. On the adoption of the terminology capitula lectionum in Wearmouth-Jarrow at the time of Bede, see M.M. Gorman, « Source Marks and Chapter Divisions in Bede’s Commentary on Luke » Revue bénédictine 112 (2002) 246-75 (p. 267); for Bede’s own use of it, see Meyvaert, « Bede’s Capitula Lectionum », pp. 348-51. 68. On the high frequency of ablative absolutes in the capitula for the Octateuch in Codex Amiatinus, see De Bruyne, « Cassiodore et l’Amiatinus », p. 264.
BBene_Boek.indbene_Boek.indb 334334 224/11/20114/11/2011 11:33:2611:33:26 H. A. G. HOUGHTON 335
43) and significare (Capp. 34, 38, 43, 45) are characteristic of this type; along with other vocabulary such as mystice (Cap. 8, cf. Cap. 5), signum (Cap. 13) and testimonium (Capp. 12, 21, 34), they reveal the compiler’s exegetical interest. This is most evident in comments which go beyond the biblical text, such as quo facto cognoscitur quod ubi ipse fuerit inuitatus uinum necesse sit deficere nuptiarum in Cap. 4 and the reference to Peter’s death by crucifixion in Cap. 45.69 The biblical text normally corresponds to the Vulgate, including dis- tinctive readings such as incredulus in John 3:36 (Cap. 7), si manseritis in sermone meo in 8:31 (Cap. 23), unum ouile in 10:16 (Cap. 26), gentiles in 12:20 (Cap. 31), confidite in 16:33 (Cap. 39) and cum fores essent clausae in 20:19 (Cap. 43). There are a handful of readings with Old Latin parallels, e.g. triduo in 2:19 (Cap. 5; cf. VL 3, 4, 14, 15), quae in deo sunt facta in 3:21 (Cap. 6; cf. VL 9A), de morte ad uitam in 5:24 (Cap. 11; cf. VL 4, 9A, 15), saturauit in a reference to 6:12 (Cap. 13, but cf. 6:26), and qui sitit in 7:37 (Cap. 19). The most interesting reading is perhaps potestatem habere se dicit ponendi ac resumendi animam suam (Cap. 27): no surviving manu- script has resumendi rather than iterum sumendi in John 10:18, although it is found in patristic citations. Nonetheless, the full integration of Vul- gate readings indicates that this series does not have Old Latin roots.70 The capitula of Type Win are an expansion based on Type C.
Type Cat. 36 chapters. Incipit (I): De diuinitate uerbi quod caro factum est, et Iohannes baptista de eo dicit: non sum dignus soluere corrigiam calciamenti eius... Explicit (XXXVI): Et cum tertio apparuisset eis piscantibus dicit Petro ter: pasce oues meas et cum senueris alius te cinget, et sequere me.
69. The comment in Cap. 4 is very similar to the Mozarabic and Spanish prolo- gues to John (ubi dominus inuitatur deficere nuptiarum uinum debeat/uinum deficiat nuptiarum). At the end of Cap. 3 (relating to Nathaniel) Codex Hubertianus includes a line from Jerome’s De uiris illustribus (et confectus senio sexagesimo et octauo anno post passionem domini mortuus iuxta eadem urbe sepultus est): this, however, seems to have been incorporated erroneously into the capitula as in its original context it refers to the evangelist John. 70. H.J. Chapman, Notes on the Early History of the Vulgate Gospels, Oxford: Clarendon Press, 1908, p. 284, suggests that the capitula in Codex Amiatinus were composed by Eugippius based on earlier divisions. While Chapman’s reconstruction of events is questionable, as there is no trace of earlier divisions matching Type C, there is nothing in Eugippius’ citations of John to contradict this; indeed, uia, ueri- tas et uita in Cap. 35 matches his non-standard quotation of John 14:6 at EUGI reg 29.93 against the form of this verse in Codex Amiatinus.
BBene_Boek.indbene_Boek.indb 335335 224/11/20114/11/2011 11:33:2611:33:26 336 REVUE BÉNÉDICTINE
This set is only found in two manuscripts, Paris, BnF, lat. 6 (tenth century) and St Millan 3 (thirteenth century), both of which are used for the text in SDR. According to the Oxford Vulgate, the series is headed incipit capitulatio, but ends expliciunt capitula. The thirty-six divisions are similar to those of Type I and there is a close but complicated rela- tionship between the titles. Many Old Latin forms are common to both, such as soluere in 1:27 (Cap. 1), natatoriam piscinam in 5:2 (Cap. 11), ire in 6:68 (Cap. 13), ante Abraham ego sum in 8:58 (Cap. 17), occidere in 12:10 (Cap. 22), flores in 12:13 (Cap. 23) and ter me negabis in 13:38 (Cap. 27); they also share the unusual editorial secessit in 4:3 (Cap. 8). Some differences between the series are attested in manuscript variants within Type I, such as the addition of ecce in Cap. 2, labores eorum in Cap. 9 and citius in Cap. 26. The last is an alteration towards the Vulgate, as is mandata mea seruate in Cap. 28 (14:16; cf. also John 4:44 in Cap. 10). These suggest that Type Cat, although longer and with more quotations, is secondary to Type I. There are also errors of sequence, consistent with the expansion of an existing series: the insertion of sicut nouit me pater et ego noui patrem from John 10:15 precedes ego sum pastor bonus (10:10) in Cap. 19; Cap. 31 ends with non tantum pro his rogo from John 17:20, but Cap. 32 (as in Type I) begins with pater sancte serua eos (17:11b). On the other hand, although some of the extra biblical text is closer to the Vulgate (e.g. scitis in Cap. 16, gentiles in Cap. 23, plorabitis et flebitis in Cap. 30), these quotations also feature a number of characteristic Old Latin forms. The most distinctive are gloriam ab inuicem quaerentes in Cap. 11 (5:44; cf. VL 8), qui non intrat per ianuam (10:1; VL 2, 3, 4, 6, 8, 13, 14, 22, 27, 35) and et ego noui patrem (10:15; VL 3, 9A, 10, 14) in Cap. 19, granum tritici in Cap. 23 (12:24; VL 2, 3, 4, 5, 6, 8, 10, 14) and omnia quae habet pater in Cap. 30 (16:15; VL 2, 5), alongside a number of omissions or variations in word order. These imply that Type Cat was compiled from an Old Latin source: further evidence for this is supplied by the alteration of John 16:2 in Cap. 30 away from the version in Type I to ueniet hora ut qui interfecerit uos arbitretur se obsequium deo praestare, a form very close to VL 8 and other members of Group 2A. Four readings are not supported in any surviving Old Latin codices, although all have parallels in patristic citations: ueniet hora quando and ipsum accipietis in Cap. 11 (John 5:28 and 5:43), nemo uenit ad me nisi pater meus traxerit eum in Cap. 12 (John 6:44) and uenient Romani et tollent regnum nostrum in Cap. 21 (John 11:48). The last two display alterations char- acteristic of flattening, suggesting that they may have been quoted from
BBene_Boek.indbene_Boek.indb 336336 224/11/20114/11/2011 11:33:2611:33:26 H. A. G. HOUGHTON 337
memory. 71 This would also explain two major lapses: the interpolation of et quadraginta annos nondum habes from 8:57 into the summary of John 2 in Cap. 5 (cp. quadraginta et sex annis aedificatum est templum hoc in 2:20; 8:57 actually reads quinquaginta), and ego sum lux mundi (John 8:12) in place of ego lux in mundum ueni (John 12:46) in Cap. 24. While the Old Latin readings are of interest, this looseness suggests that caution is nec- essary in using this series as evidence for earlier texts of the Gospel.
Type Cy. 68 chapters. Incipit (II): De Iohanne quia testimonium perhibebat de lumine, nam non erat ipse lumen, et de lumine uero et quia uerbum caro factum est. Explicit (LXVIII): Ubi apparuit discipulis tertio ad mare et prandit cum illis et dixit Petro: pasce oues meas, et de cruce eius significauit et de Iohanne dixit: sic illum uolo manere donec uenio. This series has already been studied in detail with a textual com- mentary by De Bruyne.72 The siglum Cy represents his belief that these capitula are roughly contemporary with Cyprian: De Bruyne’s preference was to date them a few years before Cyprian, accompanying a revision of the African text of the Gospels represented by Codex Bobiensis (VL 1) prior to Cyprian’s quotations and Codex Palatinus (VL 2), although he acknowledged that the reference to the adultera in Cap. 30 might post- pone them to the end of the third century.73 The capitula for John are preserved in two manuscripts, the ninth-century Vatican, Barberini lat. 637 and the tenth-century Munich, BSB Clm 6212.74 The first title is missing from both. The biblical text, with affinities to the citations of both Cyprian and Tertullian, is earlier than that of any surviving Old
71. For flattening, see H.A.G. Houghton, « ‘Flattening’ in Latin Biblical Cita- tions » in J. Baun, A. Cameron, M. Edwards and M. Vinzent (edd.) Studia Patris- tica vol. XLV, Leuven: Peeters, 2010, pp. 271-6. 72. De Bruyne, « Quelques documents nouveaux »; for John see pp. 316-24. Com- parison of the text in SDR with the extant manuscripts has brought to light a few minor omissions, corrected in the text provided in the Vetus Latina Iohannes. 73. De Bruyne, « Quelques documents nouveaux », p. 442. 74. The Munich manuscript was only discovered by De Bruyne after the publi- cation of the first installment of « Quelques documents nouveaux »; both are used in SDR. According to McGurk, its exemplar was a sixth-century manuscript from Ravenna and it has an unusual set of canon tables (« The Disposition of Numbers », note on p. 245). In addition, there are two further partial witnesses to the headings of the Cy series, Florence, BML, Edili 125 for Matthew and Paris, BnF, lat. 277 for Mark, while some of the Cy divisions are found in Matthew in London, BL, Harley 1775 (see V. Popovic ´ , « Sur l’origine de l’évangéliaire latin de la British Library, Harley 1775 » Comptes-rendus des séances de l’Académie des inscriptions et belles-lettres 134.3 (1990) 709-735, p. 728).
RB 23
BBene_Boek.indbene_Boek.indb 337337 224/11/20114/11/2011 11:33:2611:33:26 338 REVUE BÉNÉDICTINE
Latin manuscript of John. There is a consistency in the renderings (e.g. magistratus saeculi in Cap. 41 and 50 for ἄρχων κόσμου in John 12:31 and 14:30) which, along with the close sequential treatment of the text, indicates that the compiler relied on a gospel manuscript. A number of renderings are unique to these capitula, including nata est quaestio in Cap. 11 (3:25; ἐγένετο ζήτησις), saluum fecit in Cap. 18 (5:9), conprehendere in Cap. 28 (7:44), conuincit in Cap. 31 (8:46), schisma in Cap. 33 (10:19; cf. VL 2 in 7:43), constanter in Cap. 34 (10:24; cf. VL 13 in 7:13), quicumque cognosceret in Cap. 36 (11:57), claritatem in Cap. 42 (12:41), excipit in Cap. 45 (13:20), manifeste in Cap. 52 (16:29) and abstulit auriculam puero in Cap. 54 (18:10). Others are only paralleled in early quotations, such as expulit in 2:15 (Cap. 8), edidistis in 6:26 (Cap. 22), inuenit in 14:30 (Cap. 50) and officium deo facere in 16:2 (Cap. 50). As the earliest set of New Testament chapter divisions known to survive in any language, these are of considerable interest and their importance for the biblical text is unparalleled.75
Type D. 41 chapters (?). Incipit (I): In principio erat uerbum et quod in propria uenit. Explicit (XLI?; XXXIX in M; XLIII in T): ... et cum tertio manifestaret se dominus discipulis suis ait Petro: pasce oues meas et sequere me.
This series is poorly attested: no surviving manuscript has Type D capitula in all four Gospels and there are only three witnesses for John.76 In the Cutbercht Gospels (V; Vienna lat. 1224), capitula for Matthew (Type Pi) and Mark, Luke and John (Type D) are found in a group at the beginning of the manuscript, with additional sets immediately preceding Luke and John (Types A and B respectively, both entitled elenchus). The second volume of the Montpellier Bible (M; London, BL, Harley 4773) has Type A for Matthew and Type D capitula before each of the other three Gospels. Netzer’s detailed investigation of the Trier Gospels (T; Trier, Domschatz 134/61) has demonstrated that although the capitula for Matthew are Type Pi, the other three Gospels have a conflation of Type I and Type D created by the copyists from the two exemplars used
75. For an example of their application to the study of the Greek tradition, see C.-B. Amphoux, « Les premières versions latines de Luc 5 et leur contribution à l’histoire du texte » in R. Gryson (ed.), Philologia Sacra I. Altes und Neues Testa- ment. Freiburg: Herder (1993) pp. 193-211, especially pp. 200-8. 76. De Bruyne, « Quelques documents nouveaux », p. 433 identifies the Type D capitula for Matthew as a fourth-century African series, but this does not appear to apply to the other Gospels.
BBene_Boek.indbene_Boek.indb 338338 224/11/20114/11/2011 11:33:2611:33:26 H. A. G. HOUGHTON 339
for this codex. 77 (In John, this is evident from the duplication in Cap. 18, De muliere in adulterio depraehensa in moechatione.) The edition of the Type D capitula in SDR, conflating V and M with the interpolated text of T, is therefore misleading and a revised text is provided in the Vetus Latina Iohannes. The original series, represented by V and M, appears to consist of 41 chapters: in V the titles are unnumbered and not always clearly separated, while the numbering in M is confused in both the capit- ula list and the divisions in the Gospel.78 In T, these have been expanded to 43; passages unique to T are for the most part evidence for the version of Type I present in one of its exemplars.79 Even so, the text of V and M appears to be an early revision of the Type I capitula comparable to Type Cat, with extra biblical quotations and a few alterations to the chapter divisions. Some overlaps with Type Cat (e.g. the addition of ubi erat fons Iacob in Cap. 9, facta sunt encaenia in hierusolimis in Cap. 23, si quis mihi ministrat etc. in Cap. 28, pater uenit hora etc. in Cap. 37) suggests that they may be related, although Type D also has material not present in Type Cat or any other series (e.g. John 17:12 and 24 in Cap. 38). It is closer to the Vulgate than either Type I or Type Cat, as shown by readings such as si quis sitit (Cap. 17; John 7:37), antequam Abraham fieret ego sum (Cap. 20; John 8:58), qui non intrat per ostium (Cap. 22; John 10:1), cum autem tanta signa fecisset (Cap. 29; John 12:37), mansiones multae sunt (Cap. 32; John 14:2), confi- dite ego uici mundum (Cap. 36; John 16:33) in addition to cum lanternis et facibus and misit eum Annas ligatum ad Caipham (Cap. 39; John 18:3 and 24). In the light of this, the few non-Vulgate readings seem of limited sig- nificance: manducastis de panibus in Cap. 15 (6:26; VL 2, 3, 4, 5, 13, 14, 15, 22), unde ueni aut quo uado in Cap. 19 (8:14; VL 10, 11A, 47, 48), ne
77. Nancy Netzer, Cultural Interplay in the Eighth Century: The Trier Gospels and the Making of a Scriptorium at Echternach. Cambridge: CUP, 1994, especially pp. 18-21 and 162-71. SDR (p. 413) and McGurk, Latin Gospel Books (p. 114) record the Matthew capitula as absent and describe the capitula as Type I in Mark and Type D in Luke and John; in fact, the prefatory material for Matthew is misplaced, and is currently bound in the middle of John, on folios 143r to 147v. The lists for Matthew, Luke and and John have the heading breues, but Mark has capitulare. Netzer also observes (p. 21) that the additional series at the beginning of the Cutbercht Gospels may therefore come from the same scriptorium as the Trier Gospels. 78. The titles are numbered from 1 to 39, with 27 and 28 conflated and the nume- ral 27 then repeated; the divisions accompanying the text of John are given as 1-37 and 40-1. 79. There are a few places where material unique to T is not found in the Type I tradition (Capp. 1, 4, 12, 17, 18, 32, 41); these are biblical quotations which may have been added by the copyists. T is another witness to moechatione in Type I Cap. 16 and putet se officium deo facere in Type I Cap. 30 (see below).
BBene_Boek.indbene_Boek.indb 339339 224/11/20114/11/2011 11:33:2611:33:26 340 REVUE BÉNÉDICTINE
uos tenebrae conprehendant in Cap. 28 (12:35; VL 3, 11A) and uolo ut ubi ego sum et ipsi sint in Cap. 38 (17:24; VL 30) all occur in the later mixed- text tradition. The forms quid faciemus quia homo iste tanta signa facit in Cap. 25 (11:47), and unus uestrum me tradet in Cap. 31 (13:21) have no parallels in surviving manuscripts or Christian authors: they may be authorial adaptations comparable to the paraphrased form of John 7:3 in Cap. 16 (ut discipuli tui uideant quae facis).
Type I. 36 chapters. Incipit (I): Iohannes testimonium perhibet de Christo dicens: non sum dig- nus corrigiam calciamenti eius soluere. Explicit (XXXVI): Et cum tertio manifestaret se Iesus discipulis, ait Petro dicens ter pasce oues meas et sequere me.
This is the most widespread series of capitula in manuscripts with an Old Latin text of the Gospel, present in VL 8 (ff; Paris, BnF, lat. 17225, Codex Corbeiensis) and VL 6 (Paris, BnF, lat. 254, Codex Colbertinus) from Burton’s Group 2A, as well as VL 7 (G; Paris, BnF, lat. 11553, Codex Sangermanensis primus) and VL 15 (A; Stockholm, KB, A.135, Codex Aureus) from Group 2B. It also appears in the Echternach Gospels (E; Paris, BnF, lat. 9389), and manuscripts related to VL 15 including the Maaseyck Gospels (K; Maaseyck, Sint Katerinenkerk s.n.) and the Augsburg Gospels (Augsburg, Univ., Oettingen-Wallenstein’sche Bib. 1.2.4o.2; formerly Maihingen and Schloss Harburg). In addition, it occurs in an insular group comprising the Book of Armagh (D; Dublin, Tri nity College = TCD 52), the Book of Durrow (U; TCD 57) and the Book of Kells (Q; TCD 58). Thirty-two manuscripts with this series are found in SDR, although not all feature in the index on pp. 412-4 and only eleven are used for the edition of the text.80 McGurk lists twelve codices cop- ied before the year 800 with Type I in John: of these, Paris, BnF, lat. 260 uses Ammonian sections in place of chapter numbers in the capitula, while Poitiers 17 (like VL 6) has only 35 titles.81 While most manu-
80. In addition to the eight manuscripts given letters above, De Bruyne uses the following for the text: Vatican lat. 7223 (H), Paris, BnF, lat. 11957 (C) and Mont- pellier, Bibl. de la ville 3 (M). See also Popovic ´ , « Du nouveau sur les Évangiles de Split », pp. 289-91; the designation I seems to have been used to suggest an Italian origin. 81. McGurk, Latin Gospel Books, pp. 114 and 117; the reference to the final « Cap. 126 » may be an error for Cap. 226, the Ammonian section matching the last title. (For other manuscripts with this feature, see note 14 above.)
BBene_Boek.indbene_Boek.indb 340340 224/11/20114/11/2011 11:33:2711:33:27 H. A. G. HOUGHTON 341
scripts describe the list as capitula, both Poitiers 17 and the Echternach Gospels have breuis disputatio, while Codex Aureus has breues.82 As expected, Old Latin readings occur throughout the capitula, begin- ning with corrigiam calciamenti eius soluere in Cap. 1 (1:27; cf. VL 3, 4, 10, 13, 14, 15). This does not correspond to the biblical text of the Group 2A manuscripts, suggesting that the series may go back to an earlier stage. On the other hand, there are also readings more characteristic of this group such as lux in hunc mundum in Cap. 24 (12:46; VL 4, 6, 8, 9A, 11, 11A, 46, 48*), quod facis fac celerius in Cap. 26 (13:27; VL 3, 4, 8, 10) and praecepta mea custodite in Cap. 28 (14:15; VL 6, 8). The value accorded by De Bruyne to VL 8, the earliest surviving witness to the Type I capitula, is not always justified. For example, in Cap. 16 he fol- lows it by printing mulierem in adulterio deprehensam even though the six manuscripts described as β (AKCVBN) and VL 6 have mulierem deprae- hensam in moechatione. In fact, VL 8 is the only Latin gospel manuscript with moechatio in John 8:3, and this unusual rendering is far more likely to be the original form of the capitulum: it is ironic that the alteration to the capitula in VL 8 has resulted in an inconsistency with the subse- quent biblical text.83 Similarly Cap. 30 in SDR reads ueniet hora ut qui uos occiderit putet se obsequium deo facere. Although this includes a num- ber of Old Latin forms (ueniet as in VL 2, 6, 7, 9A, 10, 11, 13, 14; occiderit as in VL 2, 3, 5, 13, 14; putet as in VL 2, 5), there is no example of obsequium deo facere in Latin manuscripts of John 16:2. The alternative reading officium deo facere has an impeccable Old Latin pedigree in early African Christian writers, including Cyprian, Tyconius, Augustine and his opponent Petilianus, not to forget Type Cy above.84 Parallels for several of the Old Latin forms in these capitula are only preserved in Codex Palatinus (VL 2). These include regnum caelorum in Cap. 6 (3:5), sicut scriptum est rather than sicut dixit scriptura in Cap. 15
82. Wordsworth and White (p. 493) claim that the Book of Durrow has breuis interpretatio secundum Iohannem, but no title is visible in the facsimile. The Ech- ternach Gospels also have the archaic kata Iohannem (found in some witnesses to Type A). 83. The rare word moechatio is first attested in PS-CY sng and does not appear to be found in quotations of John 8. 84. Manuscripts with officium deo facere in Cap. 30 generally also read moechatione in Cap. 16 (e.g. VL 15, Bodley 155, the Trier Gospels and St Petersburg, NLR, Q.v.I.26). The patristic references are AU ep 185.20, AU Fau 22.70, AU Gau (seven times), CY ep 58, CY Fo 11, CY te 3.16, LUC Ath 1.19, PETI ap AU Do (three times), TY Apc 3.1; the only instance of obsequium facere is BEA Apc 5.12.7, which may represent a partial updating of Tyconius.
BBene_Boek.indbene_Boek.indb 341341 224/11/20114/11/2011 11:33:2711:33:27 342 REVUE BÉNÉDICTINE
(7:38), quarta diei rather than quadriduanus in Cap. 21 (11:39) and dile- xissetis rather than diligeretis in Cap. 29 (14:28).85 Further readings are shared with a handful of other Old Latin manuscripts, such as occidere in Cap. 22 (12:10), ter me negabis in Cap. 29 (13:38) and putet in Cap. 30 (16:2). As VL 2 is the principal witness to an African text of John, these similarities (along with the patristic support) suggest that this series is of African origin. Indeed, the match for the text of Cap. 30 in Petilianus and Tyconius would support a claim that this derives from a Donatist source (compare Types A and Ben above). One further characteristic of these capitula is their looseness.86 For example, the phrase domus orationis est domus patris mei in Cap. 5 is far closer to the sentiment expressed in Matthew 21:13 and parallels than John 2:16. Omnis propheta sine honore est in patria sua in Cap. 10 also draws on the Synoptic phrasing (Matt. 13:57, Mark 6:4) instead of John 4:44.87 Cap. 29 reverses the order of verses, quoting John 15:1 before John 14:28. In John 3:5, almost all Latin gospel manuscripts have non potest with the infinitive: non intrabit in Cap. 6 is a flattened form found in numerous Christian authors which also has Synoptic parallels (e.g. Matt. 7:21, 19:23).88 An even more interesting detail is found in Cap. 11, where the colonnade around the Pool of Bethesda in John 5:2 is identified with the porticus Salomonis of John 10:23. This connection is also made by four Church Fathers: Hilary of Poitiers, Paulinus of Nola, Augustine (three times), and Cassiodorus (probably following Augustine).89 The sur- prising conclusion is that the Type I capitula were composed with com-
85. The first and last of these are not adopted in the editorial text given in SDR, but their biblical affiliation offers a strong case for their authenticity. 86. For looseness as a feature of Donatist capitula, see Bogaert, « Les particula- rités éditoriales », pp. 14-15. 87. It is possible too that secessit in Cap. 8 (referring to John 4:3) draws on Synop- tic parallels (e.g. Matt. 4:12, 14:13, 15:21; Mark 3:7; Luke 9:10): this verb is not found in John apart from certain Old Latin manuscripts in John 6:15. 88. Fischer, Die lateinischen Evangelien, p. 53 records one manuscript from the first millennium with intrabit in place of potest introire (Wb; VL 271, a liturgical Liber misticus). Tertullian, Chromatius of Aquileia, Filastrius of Brescia, Jerome, Augustine, the Council of Carthage in 411 and the early African De trinitate (ascrib ed to Vigilius of Thapsus) are among the early sources for intrabit. 89. HIL Ps 91.5; PAU-N carm 28.307; AU Jo 20.2.3; AU Ps 83.10; AU s 272B.4; CAr Ps 25.12. On this possible identification in Possidius’ list of Augustine’s sermons, see further David F. Wright, « Piscina Siloa or Piscina Salomonis? (Possidius, Indi- culum X 6.57) » Revue des études augustiniennes 25 (1979) 47-60; Wright observes that it is unlikely that Augustine influenced the capitula (p. 55), but the antiquity now demonstrated for this series means that they may have been familiar to Augustine.
BBene_Boek.indbene_Boek.indb 342342 224/11/20114/11/2011 11:33:2711:33:27 H. A. G. HOUGHTON 343
paratively little reference to the text of the Gospel and may have drawn on (or subsequently influenced) exegetical tradition.90
Type Ifor. 39 chapters. Incipit (I): De principio euangelii. Explicit (XXXVIIII): Ubi tertio se manifestauit discipulis ad mare Tibe- riadis et manducauit cum illis. In Codex Forojuliensis (CLA 285, now divided between Cividale, Prague and Venice), John is preceded first by Type B capitula described as breues and then by this unique series under the heading of capitula- tiones.91 Within the 39 numbered titles, there are 13 more subdivisions identified by ekthesis in SDR. Almost all the numbered titles begin with de or ubi, indicating that this is a unified system. The series has several of similarity with other types, as in the addition of et ementes in Cap. 4 (2:15, cf. Matt. 21:12 and parallels; also Types D, Pi, W) and the use of the non-biblical increpare for Jesus’ rebuke of Thomas in Cap. 38 (20:27; also Types Cat and I). A number are shared with Cy alone, such as the beginning of a new chapter at John 11:53 (Cy 36 and Ifor 25), the quota- tion of John 12:36 (abscondit se; Ifor 28 and Cy 41) and the term aduocatus in 14:16 (Ifor 31 and Cy 49); John 14:8 only features in Cy (49), Ben (37) and Ifor (30), while John 15:14 (amicos) is limited to Cy (50), D (34) and Ifor (32). However, there is also much biblical material which is unique to Ifor, including references to John 5:30 (Cap. 12) and 17:15 (Cap. 35) and numerous details (e.g. lucerna in Cap. 12 (5:45), Capharnaum (6:24) and qui de caelo descendi (6:48) in Cap. 14, manducauit cum illis (cf. 21:13) in Cap. 39. Some of these have Old Latin characteristics, including de supe- rioribus sum in Cap. 19 (8:23; VL 3, 5, 8, 13), occidere in Cap. 25 (11:53; VL 2, 3, 14) and both ubi honorificat patrem (17:1-5; cf. VL 3, 5, 13, 14) and conserues a malo (17:15: VL 15, cf. 10, 13) in Cap. 35; of particular interest is plurimi discipuli in Cap. 15 (6:66), only matched by Jerome, Epistula 40.1.3. These indicate that Ifor is not dependent on another series for common Old Latin renderings and should therefore be treated as a further independent set of Old Latin capitula.92
90. It is conceivable that the Synoptic parallels may derive from a list of Ammo- nian sections with titles similar to those found in VL 10, 39, 40 and 46: in Canons comprising two or more Gospels, the Matthaean (or Synoptic) text is usually quoted, although this does not explain all the inaccuracies. 91. For other uses of this term, see Petitmengin, « Capitula païens et chrétiens », p. 495 and Type Cat. 92. Old Latin forms shared with other series include natatoria in 5:2 (Cap. 10, cf. Types A, B, Cat, I), ianua in 10:1-7 (Cap. 22, cf. Types A, B, Ben, Cat), dedicatione
BBene_Boek.indbene_Boek.indb 343343 224/11/20114/11/2011 11:33:2711:33:27 344 REVUE BÉNÉDICTINE
Type In. 36 chapters. Incipit (I): Iohannes testimonium perhibet de ipso et clamat dicens. Explicit (XXXVI): Postea manifestauit se iterum Iesus ad mare Tiberia- dis.
This series is only present in the sixth-century Burchard Gospels (Würzburg, Univ., M.p.th.f. 68) two later copies (Würzburg, Univ., M.p.th.f. 65 and 66). The divisions are identical to Type I as found in Codex Aureus (VL 15), a close relation to the Burchard Gospels in John, but the titles have been replaced by the opening words of each chapter. Their biblical text is typical of an early Vulgate witness with a sprinkling of Old Latin features, such as die tertia in 2:1 (Cap. 2), the omission of eis from 8:21 (Cap. 17), altera die in 12:12 (Cap. 23) and turbatus est in spiritu in 13:21 (Cap. 26): all these are paralleled by the subsequent text of the Gospel. An even more compelling demonstration of this connection is Cap. 11, beginning et ascendit Iesus: the first half of John 5:1 is also omitted by the first hand. However, there remain six minor discrepan- cies between the capitula and the Gospel, which may be explained as misreadings of the exemplar or subsequent alterations to its text prior to copying.93
Type P i. 35 chapters. Incipit (I): Ubi Iohannes testimonium perhibet de Christo. Explicit (XXXV): Ubi dominus resurgens ianuis clausis discipulis appa- ruit. item post dies octo similiter tertioque ad mare Tiberiadis piscantibus apostolis se praebuit ac Petro dicit pasce oues meas.
SDR lists thirty-two manuscripts with Type Pi in John, of which five are used for the edition: London, BL, Add. 10546 (K); London, BL, Harley 2790 (H); Nancy, Cathedral s.n. (N); Bamberg, A.I.5 (B); Paris, BnF, lat. 13169 (P). The last is VL 29, a mixed-text manuscript which only has capitula before John and no chapter divisions in the text: it introduces the series as capitula parabolarum. The sole pre-ninth-century
in 10:22 (Cap. 23, cf. Types A, B), flores in 12:13 (Cap. 26, cf. Types Cat, D, I), Graeci in 12:20 (Cap. 27; cf. Types A, B, Cy), tritici in 12:24 (Cap. 27, cf. Types Cat, Cy), uinea in 15:1 (Cap. 32, cf. Types A, B) and amat in 16:27 (Cap. 34, cf. Type C). 93. These are as follows: in Hierosolymis in Cap. 11 but Hierosolymam (first hand; corrector Hierosolymis) in 5:1; uidens in Cap. 2 but uidet in 1:29; consequentes in Cap. 3 but eos sequentes at 1:38; colligerunt autem in Cap. 22 but colligerunt ergo in 11:47 (where there is no VL parallel for autem); diligitis and seruaretis in Cap. 28 but dili- geritis (first hand; corrector diligitis) and seruate at 14:15 (where, again, seruaretis is not found in Old Latin manuscripts).
BBene_Boek.indbene_Boek.indb 344344 224/11/20114/11/2011 11:33:2711:33:27 H. A. G. HOUGHTON 345
witness in McGurk is the Livinus Gospels (Ghent, St Bavo 13), appar- ently not used by De Bruyne, which has the heading capitula and lacks the final three titles. The dependence of Type W on this series, how- ever, would push its composition back to before the sixth century. Paris, Arsenal, 1184 divides the final title into four, giving a total of thirty- eight headings. The chapter divisions and text of the series often overlap with Type I: it appears to represent an intermediate stage between Type I and Type W, abbreviating the Type I capitula, frequently adding the phrase ubi dicit and expanding some of the biblical quotations. Very few of the non- Vulgate forms from Type I remain, which suggests that the capitula have been compared with the biblical text. Cap. 13, for example, reads multi discipulorum instead of aliqui de discipulis (6:66). In Cap. 23 ramos pal- marum replaces flores palmarum and material is cited from the rest of John 12:13; the same is true of the next title, with tanta signa for multa signa followed by the subsequent part of 12:37. Likewise, the form of John 14:15-16 in Cap. 27 matches later versions (cp. Type I Cap. 28). Most of the additional biblical material corresponds to the Vulgate, with a handful of minor exceptions. These include pellit in Cap. 5 (in place of eiecit at 2:15) and saluat in Cap. 9 (the Vulgate reads sanaret at 4:47), both of which may be loose references rather than quotations, and perhi- bet rather than perhibebit in Cap. 30 (15:26). The most striking is ianuis clausis in the long final capitulum, apparently pertaining to 20:19 (where it is only attested in VL 2, cf. VL 27). However, as this phrase occurs in the Vulgate at 20:26 it is possible that the compiler took it from there (cf. Types D and Vich).
Type Vich. 13 chapters. Incipit (I): De diuinitate uerbi dei et Iohannis missione ... atque angelos super se descendentes et ascendentes uisuros praedicit. Explicit (XIII): Iesus flagellatur, spinis coronatur, uestitur purpura, alapis caeditur ... Iohannem in pace quieturum designat.
This series only occurs in two manuscripts, both held in the Town Museum in Vic in Catalonia: the eleventh-century MS 89, and its twelfth- century copy, MS 119. The divisions are similar to the fourteen chapters of Types A and B, with chapters 11 and 12 combined. Similar opening lines for Capp. 2, 3, 6 and 11 might suggest that one of these was taken as a model, but this lengthy series seems to be a new composition for an existing set of divisions, citing verses not present in any other set of
BBene_Boek.indbene_Boek.indb 345345 224/11/20114/11/2011 11:33:2711:33:27 346 REVUE BÉNÉDICTINE
capitula (e.g. 1:51, 5:16, 7:6, 7:40-1, 10:39 etc.). The biblical text is for the most part consistent with the Vulgate, including characteristic read- ings such as probatica piscina in John 5:2 (Cap. 5), ostium in 10:7 (Cap. 9), gentibus in 12:20 (Cap. 10) and alapam in 18:22 and 19:3 (Capp. 12, 13). Differences such as saluificet in Cap. 2 (cf. 3:17), uero in Cap. 3 (cf. 3:30), iuxta puteum in Cap. 4 (4:6) and comederint in Cap. 6 (6:26) may be ascribed to the heavy redactional activity of the compiler, although crucem portat (cf. 19:17; VL 10, 13, 14) and ianuis clausis (20:19; cf. Types D and Pi above) find some support elsewhere.
Type W. 35 chapters. Incipit (I): Testimonium Iohannis de Christo. Explicit (XXXV): Resurrectio domini.
Type W is peculiar to the sixth-century Kilian Gospels (Würzburg, Univ., M.p.th.q. 1a), occurring with the heading breues before Mark, Luke and John (the beginning of Matthew is not extant). It appears to be dependent on Type Pi: the scope of every capitulum for John is an exact match and five titles are identical (Capp. 8, 11, 16, 18 and 34). On the other hand, the biblical text is accommodated to the Vulgate three times (eicit in Cap. 5, addition of autem in Capp. 24 and 30), while direct quotations replace the summaries of Type Pi in Capp. 23, 28 and 31. This shows that the adaptation of an existing series could be combined with the introduction of new material.
Type Win. 42 (unnumbered) chapters. Incipit: Ubi in principio uerbum esse et apud deum esse et deus esse. euan- gelizante Iohanne memoratur ... omnes in se credentes facit filios dei fieri per gratiam suam. Explicit: Ubi dominus Petro pascendas oues aeque tertio commendans. ... et de conclusione euangelistae quod uerum sit testimonium eius.
This is the only set of titles for John listed by Berger which is not included in SDR.94 It is found in a single manuscript, Oxford, Bod- leian L., Auct. E. inf. 2, the second part of a two-volume Bible copied in Winchester in the twelfth century. Although the capitula before Matthew correspond to Type Pi, those for the other three Gospels all seem to be expanded versions of earlier series. The source for John is clearly Type C although the division of titles (marked by large capital letters) is not
94. Berger, Histoire de la Vulgate, p. 355, Type I.13.
BBene_Boek.indbene_Boek.indb 346346 224/11/20114/11/2011 11:33:2711:33:27 H. A. G. HOUGHTON 347
identical. In addition to distinctive phrases such as excitandi mysterium corporis sui proponit (Cap. 5), mulieri Samaritanae plurima mystice loqui- tur (Cap. 8), pharisaeos de proprio testimonio reluctantes arguit (Cap. 21), hora clarificationis suae praedicit (Cap. 30) and prolixa et multimoda ora- tione commendat (Cap. 38), there is considerable verbal overlap through- out. That this is secondary to Type C is indicated by the replacement of pronouns, the omission of Cap. 12 and substitution by Vulgate forms, as in sub condicione qua amplius non peccaret (Cap. 20; Type C reads ulte- rius). The additional biblical material (e.g. the extracts from John 19 in the Passion narrative) also corresponds to the Vulgate.
Type Z. 40 chapters. Incipit (I): De incarnatione uerbi et testimonio Iohannis. Explicit (XL): Ubi dominus interrogat Petrum si diligat eum et ubi dixit ei sequere me.
This series is known only from one twelfth-century gospel book with glosses, Oxford, Bodleian L., Laud lat. 25. The nature of the capitula in this manuscript is more complicated than is apparent from SDR. On fol. 11v there are partial and disordered lists of titles quoting verses from each Gospel. These are followed by two complete sets of capitula for Mat- thew and John, apparently in the same hand, which are the sole instances of Type Z.95 All the titles begin with ubi or de and usually quote biblical material from the beginning of each chapter. The gospel text matches the Vulgate, with probatica piscina in 5:2 (Cap. 12), ostium in 10:1 (Cap. 22), gentiles in 12:20 (Cap. 28) and palmites in 15:5 (Cap. 32): even super puteum in 4:6 (Cap. 10), the single possible exception, has some currency in Vulgate manuscripts. As there is no match with any other series of divisions for John, it seems best to describe this series as a one-off based on the Vulgate.96
Conclusion This survey has demonstrated that a surprising number of series of capitula were composed with reference to pre-Vulgate versions of John
95. A collation with the text in SDR provides the following emendations: Cap. 9, discipulorum corr., added above Iesu; Cap. 12, ierosolamis] ierosolimis; Cap. 28, iesum] iesu; Cap. 31,
BBene_Boek.indbene_Boek.indb 347347 224/11/20114/11/2011 11:33:2811:33:28 348 REVUE BÉNÉDICTINE
and continued to be transmitted in manuscripts with Jerome’s revised text of the Gospel. The most ancient is the remarkable African Type Cy, dated by De Bruyne to the middle of the third century. As observed above, this is the earliest known series of gospel chapters in any lan- guage. Type I is likely to be the next oldest set: the early readings such as moechatione in John 8:3 and officium deo facere in 16:2 along with similarities to the text of Codex Palatinus suggest that it was composed in Africa in the fourth century. It is possible that its compilation was undertaken in conjunction with the revision of the gospel text underly- ing Group 2A, although witnesses to the latter are generally closer to the Vulgate. Numerous revisions and reworkings are evident both in the variations within Type I and the headings of other thirty-six (and thirty- five) chapter series. Of these, Type Cat and the original version of Type D incorporate additional material from an Old Latin version of the text, while Type Pi comes from the early period of the Vulgate and Type W is an abbreviation of Type Pi not later than the sixth century. The other major Old Latin series is that of the fourteen-chapter Type A (formerly known as Br). Its correspondence with the early form of text found in VL 14 (matching VL 4 and VL 47 in the early chapters of John) again points to a fourth-century origin. It was suggested above that the choice of topics in these capitula might indicate a Donatist origin, match- ing the series identified for other biblical books. Type I, however, with its more marked African characteristics, has an equally strong contention to be a product of a Donatist milieu. Type B (formerly known as B=A) is a shorter form of Type A, abbreviated at a fairly early stage but pre- serving the form of the biblical quotations. In addition, it appears that Type Ben and Type Ifor (the latter only found in Codex Forojuliensis) are independent compilations with some Old Latin characteristics. The remaining series are of Vulgate origin. Type In, peculiar to the Burchard Gospels, was produced in the fifth or early sixth century: it replaces Type I with the opening words of each chapter, possibly taken from the exemplar used for the subsequent text of John. Type C, the capitula with the distinctive heading capitula lectionum found in Codex Amiatinus and several insular manuscripts, is a new series based on a Vul- gate text; Type Win is a later expansion of these. Type Vich and Type Z are very rare and are not attested before the eleventh and twelfth cen- turies respectively: their similarity to earlier series of divisions suggests that they were composed to supply or replace lists of headings for a text of the Gospel which already had numbered divisions.
BBene_Boek.indbene_Boek.indb 348348 224/11/20114/11/2011 11:33:2811:33:28 H. A. G. HOUGHTON 349
As well as providing important evidence for the history of the Latin versions of the text of John, these series of capitula are also significant for studying the production of gospel codices and the use and interpre- tation of the Bible. The fact that the most archaic capitula (Type Cy) are only preserved in two Vulgate manuscripts from the ninth and tenth centuries, with all witnesses to Type A of a similar date, demonstrates the eclectic nature of the prefatory material in gospel books. The trans- mission of these ancient forms and, indeed, the difference between most series and the subsequent text of the Gospel suggests that this discrep- ancy was not often noted, and one may wonder how frequently the capi- tula (or, indeed, the chapter numbers) were used as a system of reference: it is also not unusual for the numeration of the titles to bear no relation- ship to the divisions alongside the following text of the Gospel. The num- ber of different series produced between the fourth and sixth centuries is remarkable: perhaps it reflects a growing emphasis on the form of the scriptural text along with the fixing of the canon. It is noteworthy that this activity is confined to the Latin tradition, and there is no overlap with Greek systems.97 It is also interesting to speculate on the extent to which textual features of the capitula, such as the conflations in Type I, may have influenced the form in which Christian authors quoted or expounded the gospel.98 De Bruyne remarks that the creation of each new set of capitula represented a new edition of the biblical text, while Berger asserts that the study of the summaries is indispensable for the history of the biblical text.99 I hope that this survey building on their valuable work will lead to this evidence being given renewed attention and that the new edition of the Vetus Latina Iohannes will provide a sound basis for fresh research on the Latin versions of John.
University of Birmingham, Hugh A. G. HOUGHTON Institute for Textual Scholarship and Electronic Editing ([email protected])
97. On the sets of Latin capitula for Hebrews and the Song of Songs which are related to Greek series, see Bogaert, « Les particularités éditoriales », p. 8. 98. See also P. Petitmengin, « Les plus anciens manuscrits de la Bible latine » in J. Fontaine and C. Pietri (edd.) Le monde latin antique et la Bible (Bible de tous les temps 2), 89-128, especially pp. 102-3. 99. De Bruyne, « Quelques documents nouveaux », p. 295; cf. De Bruyne, « Cas- siodore et l’Amiatinus », p. 263; Berger, Histoire de la Vulgate, p. 315.
BBene_Boek.indbene_Boek.indb 349349 224/11/20114/11/2011 11:33:2811:33:28 350 REVUE BÉNÉDICTINE
APPENDIX
TABLE OF DIVISIONS IN CAPITULA SERIES AND OLD LATIN MANUSCRIPTS
This table is based on that in Sommaires, divisions et rubriques de la Bible latine (pp. 521-6) and the transcriptions of Old Latin manuscripts in the Vetus Latina Iohannes Electronic Edition. Several alterations have been made to the earlier table, including the correction of misprints and the comparison of divisions with the titles of each series in order to arrive at an archetype for each. It should be remembered that the signalling of divisions in gospel manuscripts often fails to correspond exactly to the archetypical divisions, especially in the more widely-attested series.
KEY R = rubrication (no number present) C = double-height capital (no number present) CC = outsize capital (no number present) ( ) = division indicated in manuscript (e.g. capitals, rubrication) but no number visible italic font = division reconstructed from title X = superscript letter indicates manuscript with variant placing of division
= verse absent from manuscript
= manuscript no longer extant
BBene_Boek.indbene_Boek.indb 350350 224/11/20114/11/2011 11:33:2811:33:28 [1] [2] [4] [3] Γ Δ Β Α C C C CC
(8) (5) ). 1 capitula 55 5 3 3 3 (3) 3 tching the [4] 4 8 8 8 8 8 [3] 3 7 (7) 7 7 (7) 7 C Γ Δ 5 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 2 8 1R CC Z 5 3 810 P A.1 (O). for 6 5 1 Rouen 4 3 4 5 2 R 3 2 2 4 4 4 (4) 3 ! C one. 5 654 6 3 5 C 4 11 2 5 Ben C I 2 6 8 877 7 9 5R43 2111 433 4 1R C Reg. lat. 14 (R) and 56 6 7 7 6 6 7 4 R 6 (6) 6 6 6 C 45 2 3 2 2 3 2 (2) 2 (2) 2 C 8 9 988 6R 11 Cat D 2 Vatican 7 5 1 6 3 4 2 W i3 H 11 4 4 titles as well divisions in Harley 2790 (H) and Arsenal 1184 (A). series (in groups; revised and supplemented) Vetus Latina manuscripts (in groups) Q I In P (6) 8 8 8 (5) 7 7 7 (1) (3) 4 4 4 and are based principally on
SDR 2 capitula SDR 4 3 1 2 B Vich titles (Types Vich and W) or supplied from manuscripts D Z). 1 5 in Type Ben, and only 49 total (despite the 50 capitula). All subsequent numbers have therefore been increased by one (ma capitula Cy A capitulum ; the divisions are not included in r lium fi : reconstructed from in Type I appears to quote John 1:15, which may have led confusion over the placing of this division (compare VL 6, 7, 15). SDR capitulum divisions have been revised based on the extent of i Verse Incipit (Vulgate) 4:1 ut ergo cognouit 12 4 4 3:35 pater diligit 3:243:25 nondum enim missus facta est ergo quaestio 11 7 3:133:163:22 et nemo ascendit sic enim dilexit deus post haec uenit 3 3 2:122:13 post hoc descendit et prope erat pascha 8 5 5 5 1:191:26 et hoc est testimonium respondit eis Iohannes 3 1 1 1 2 1 2 2 1 2:13b2:14 et ascendit Hierosol. 2:233:1 et inuenit in templo cum autem esset erat autem homo 9 10 (4) 6 6 6 1:1 in principio erat 1 1:351:38 altera die iterum conuersus autem 5 (2) 3 3 3 3 1:281:29 haec in Bethania altera die uidet1:401:411:43 erat autem Andreas 4 2:1 inuenit hic primum in crastinum uoluit et die tertio nuptiae 6 7 2 2 2 2 2 4:34:44:5 reliquit Iudaeam oportebat autem uenit ergo in ciuitat. 1:61:91:15 fuit homo missus erat lux uera Iohannes testimonium 2 2 Not present in 3 Type P 1 Formerly known as Type B 4 The first 5 SDR has no division corresponding to 6 The division at 2:13 is missing from Harley 4773 (M); occurs 3:1 and all subsequent divisions are decreased by
BBene_Boek.indbene_Boek.indb 351351 224/11/20114/11/2011 11:33:2811:33:28 BBene_Boek.indb 352 e n e _ B o e k . i n d b
3 5 2 Cy A B Vich Q I In Pi W Cat D Ben C Ifor P Z 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 4:6 erat autem ibi fons 13 (9) 4:6b Iesus ergo fatigatus 9 9 8A 8? 910 9999 9 4:7 uenit mulier E [5] 4:28b et dicit illis homin. 14 4:40 cum uenissent ergo 15 4:43 post duos autem dies (7) 10 10 9A 9 10 11T 9 7 R 10 10 10 10 (10) 10 C 4:44 ipse enim Iesus 16 9H 10 4:46 uenit ergo iterum Ϛ [6] 4:46b et erat quidam regulus 17 [11]M 911 4:54 hoc iterum secund. 5R 5 5:1 post haec erat dies 5O 5 5 (8) 11 10 10 11 12 11 10 10 9 12 8 R 6 5 11 (11) 11 11 (11) 11 C 5:1b et ascendit Iesus 11 5:5 erat autem quidam Ζ [7] 5:9b erat autem sabbatum 18 5:14 postea inuenit eum 19 5:18b respondit itaque Iesus 13 12 11 5:21 sicut enim pater 11 5:25 amen amen dico uo. (9) 9R 5:28 nolite mirari hoc 13 5:30 non possum ego 12 5:31 si ego testimonium 12 C 6:1 post haec abiit 6R (10) 12 12 11 11 12 14 13 13 10 13 (10) R 6 Ε [5] 6 12 (12) 12 (12) 12 C 6:3 subiit ergo in mont. 6 6 6 14 6:4 erat autem proxim. 7 6:5 cum subleuasset ergo Η [8] 6:7 respondit ei Philippus 20 6:15 Iesus ergo cum cog. 21 11 6 6:15b fugit iterum in mont. C 6:16 ut autem sero 12 12 15 6:19 cum remigassent ergo Θ [9] 6:22 altera die turba (11) 15 14 14 11 R 6:26b amen amen dico uobis 22 6:35 dixit autem eis Iesus 16 15 6:41 murmurabant ergo 16 6:44 nemo potest uenire 14 6:47 amen amen dico uobis 12 6:48 ego sum panis uitae 13A7 13 224/11/2011 11:33:29
4 7 There is no division 13 for Pi in SDR or Harley 2790 (H). / 1 1 / 2 0 1 1
1 1 : 3 3 : 2 9 [10] Ι C C
(18) 16 13 13 13 [7] 7 14 (14) 14 (14) 14 C 12 C 14 Ζ 16 18 : the table has been adjusted to match these. 22 17 R 18 18 (18) 18 R 18 17 2021 16 15 capitula P Z 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 12 19 (16) 8 for 21 17 14 20 G 19 23 17 20 26 26 18 11 24 24 17 19 22 21 15 17 19 18 15 R 15 15 15 (15) 15 18 21 25 25 16 18 21 20 19 18 16 16 16 14 16 18 16 16 (13) R 8 9 17 13 17 15 18 16 14 W Cat D Ben C I : all manuscripts investigated place it at 8:21. i SDR (12) 14 14 14 (14) 17 17 20 23 22 R 9 15 8 17 (17) 17 (17) 17 7 8 (20 at 8:12, 21 8:31, 22 9:1, 23 10:1, 24 10:22) do not correspond to the for 9 9 for I SDR B M Cy A B Vich Q I In P 7:457:46 uenerunt ergo responderunt ministri 29 9:39 dixit ei Iesus in iudic. 8:38:12 adducunt autem scribae iterum ergo locutus 30 17 7:27:7b7:9 erat autem in prox. 7:11 quia ego testimonium7:14 haec cum dixisset 7:15 Iudaei ergo quaerebant 24 7:16 iam autem die festo 7:30 et mirabantur Iudaei 25 7:31 respondit eis Iesus 7:32 quaerebant ergo eum 26 7:37 de turba autem multi audierunt pharisaei 27 in nouissimo autem 28 (13) 15 15 15 8:12b ego sum lux mundi 8:468:518:52 quis ex uobis arguit amen dico uo. dixerunt ergo Iudaei 8:208:21 haec uerba locutus est 8:30 dixit ergo iterum eis8:31 haec illo loquente dicebat ergo Iesus 8:59b 8 31 9:1 et exiuit de templo 8 et praeteriens uidit 32 (15) 18 18 18 7:478:1 responderunt ergo Iesus autem perrexit 29 16 16 16 9:39b ut qui non uident 33 6:637:1 spiritus est qui post haec ambulabat 7 7 6:516:596:60 ego sum panis uiuus haec dixit in synagoga multi ergo audientes 23 13 13 138:22 dicebant ergo 17 (13) (13) 9 The division at 8:22 in Type B (which is not a logical break) seems to be an error 8 Between John and 11, the divisions given in
BBene_Boek.indbene_Boek.indb 353353 224/11/20114/11/2011 11:33:2911:33:29 BBene_Boek.indb 354 e n e _ B o e k . i n d b
3 i for 5 Cy A B Vich Q I In P W Cat D Ben C I P Z 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 4 9:41 dixit eis Iesus 9 10:1 amen amen dico uo. 9 9 (16) 19H 1910 22T 26 22 14 R 10 9 CC 10:7 dixit ergo eis iterum 19 19 19A 19 10 19 (19) 19 (19) 19 10:10b ego ueni ut uitam 27 10:11 ego sum pastor bonus 22M11 23 10:17 propterea me pater 27 10:22 facta sunt autem 34 20 20 20 20 20 23 28 23 24 20 20 20 20? 20 10:25b opera quae ego facio 11:1 erat autem quidam 35 12 21 21 21 21 21 24 29 28 24 25 21 21 21 21? 21 C ΙΑ [11] 11:37 quidam autem dixer. 36 11:47 collegerunt ergo 22 22 25 30 29 25 26 22 (22) 22 (22) 22 11:55 proximum autem erat 10 1012 10 22H 22 31 10 C 12:1 Iesus ergo ante sex 37M12 22A 22 26T 27 11 12:2 fecerunt autem ei 37B 26M 15 ΙΒ [12] 12:4 dicit ergo unus ΙΓ [13] 12:12 in crastinum autem 38 23 23 23 23 23 27 32 30 26 16 28 23 (23) 23 23? 23 C 12:14 et inuenit Iesus 39 ΙΔ [14] 12:18 propterea et obuiam 40 12:20 erant autem gentiles 41 31 27 ΙΕ [15] 12:24 amen amen dico uobis 33 12:26 si quis mihi ministrat 28T13 12:27 nunc anima mea 32 12:36b haec locutus est 28 12:37 cum autem tanta signa 42 24 24 24 24 29 24 24 24 24? 24 12:42 uerumtamen et ex 43 12:44 Iesus autem clamauit 34 12:46 ego lux in mundum 44 24 13:1 ante diem autem 11 11 11 (17+) 25H 30 35 33 29 17 29 R 11 11 25 R CC 13:2 et cena facta cum 25 25 25A 25 25 25 25 25 13:3 sciens quia omnia 45 13:4 surgit a cena ΙϚ [16] 13:16 amen amen dico uobis 34 13:21 cum haec dixisset 26 26 26 31 36 26 26 26 (26) 26 13:23 erat ergo recumbens 46 13:27 et post buccellam 18
10 In SDR division 19 in Cat is located at 10:7, but the capitulum quotes 10:1. 11 As capitulum 22 in D quotes 10:1, the division at 10:11 in Harley 4773 (M) is erroneous. 224/11/2011 11:33:30
4 12 In SDR division 10 in B is located at 11:1, but the division of the capitula suggests it should be at 11:55 (matching Type A and the majority of manuscripts investigated) or possibly 12:1 (also well attested in manuscripts). / 1
1 13 This division (and the corresponding title) are missing from Harley 4773 (M); subsequent divisions in this manuscript are now 2 below the expected number. / 2 0 1 1
1 1 : 3 3 : 3 0 [17] ΙΖ C
30 (30) 30 30? 30 12 37 36 33 P Z 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 21 for 31 33 34 16 35 30 39 37 41 40 37 40 35 R 13 31 (31) 31 31 31 35 34 39 37 [32] 19 32 R 12 29 (29) 2939 42 (29) 29 41 36 38 R 13 33 (33) (33) 33 CC 36 34 15 15 32 38 20 32 (32) 32 32 32 28 33 38 36 31 R 28 28 28 (28) 28 31 27 32 30 27 (27) 27 27 27 33 30 32 27 31 26 30 W Cat D Ben C I 28 29 29 33 i H 18 H14 19 refers to 18:13 (where Arsenal 1184 has another capital letter). suggests that it comes later (18:28 in the majority of manuscripts). capitulum (22+) 31 31 31 (23+) 33 33 capitula quotes 15:1 and 4. 12 capitulum 17 12 (19+) 29 29 29 12 Cy A B Vich Q I In P is located at 18:1 (as in Harley 2790), but the i is missing from Arsenal 1184 (A), and the remaining four numbered divisions are one short of expected numeral. i lioli adhuc modicum 27 27 26 division 38 in C is duplicated at 15:26. division 14 in B is located at 18:1, but the of division 29 in Cat is located at 15:11, but the division 33 in P fi SDR SDR SDR SDR 18:1 haec cum dixisset 14:2214:27 dicit ei Iudas pacem relinquo uobis 50 28 17:1b17:1c et subleuatis oculis 17:11b pater uenit hora pater sancte serua eos 32 32 32 13:3614:1 dicit ei Simon Petrus14:514:8 non turbetur cor 14:12b dicit ei Thomas 47 14:15 dicit ei Philippus amen dico uobis si diligitis me 48 49 (18+) 28 28 27 15:11 haec locutus sum 18:318:1018:12 Iudas ergo cum Simon ergo Petrus18:13 cohors ergo et trib. et adduxerunt eum 54 55 33 16:116:2b16:15 haec locutus sum uob. sed uenit hora 16:16 omnia quaecumque 16:19b modicum et iam non16:23b et dixit eis de hoc16:25 amen dico uobis 16:29b haec in prouerbiis 16:31 ecce nunc palam16:33 51 respondit eis Iesus 17:1 haec locutus sum uob. haec locutus est Iesus 52 53 30 (21+) 13 30 13 35 R 13:33 15:1215:17 hoc est praeceptum 15:26 haec mando uobis cum autem uenerit (20+) 30 38 R 15:1 ego sum uitis uera 16 In 17 In 15 In 14 Division 28 for P 18 In
BBene_Boek.indbene_Boek.indb 355355 224/11/20114/11/2011 11:33:3111:33:31 BBene_Boek.indb 356 e n e _ B o e k . i n d b
3 5 6
Cy A B Vich Q I In Pi W Cat D Ben C Ifor P Z 10 13 9A 11 22 47 6 7 8 11A 14 15 48 27 18:15 sequebatur autem 43 18:18 stabant autem serui 56 18:28 adducunt ergo Iesum 14 14 (24+) 34H 34 44 R 14 14 19:1 tunc ergo adprehendit 57 13 45 19:16b susceperunt autem 58 34 34 46 34 34 34 34 34 19:17 et baiulans sibi 34 19:23 milites ergo cum 59 (25+) 47 R 19:25 stabant autem iuxta 34 19:26 cum uidisset ergo 60 40T20 19:26b dicit matri suae 35 19:30 cum ergo accepisset 61 19:31b rogauerunt Pilatum 62 19:38 post haec autem 48 C ΙΗ [18] 19:40 acceperunt ergo 63 20:1 una autem sabbati 21 (26+) 35 35 35H 35 35 41T20 49 42 37 39 R 35 35 (35) 35 CC 20:16b Maria conuersa illa 64 20:19 cum esset ergo sero 65 40M22 43 38 ( ) ““ 42T* 20:24 Thomas autem unus 66 20:26 et post dies octo 41M 20:28 respondit Thomas 67 20:30 multa quidem et alia 36 21:1 postea manifestauit se (27+) 36 36 36 43T20 50 44 [39] R 36 C 21:12b et nemo audebat 68 21:14 hoc iam tertio (36) 21:15 cum ergo prandissent 45 40
19 A division in Pi at 18:13 is supported by the capitulum and Arsenal 1184; see the note above on 18:1. 20 The Trier Gospels are lacunose from 19:9 to 21:20; divisions 40–43 have been reconstructed from the capitula. 21 The capitula in Rouen A.1 suppose a fifteenth division at this point, not present in the other witnesses or Type B. 22 There are no divisions between 18:2 and 19:42 in Harley 4773 (M); they resume at 20:1 with the expected numeral 40. 224/11/2011 11:33:31 4 / 1 1 / 2 0 1 1
1 1 : 3 3 : 3 1